Opinion for the Court filed by Senior Circuit Judge BAZELON.
Opinion concurring in part and dissenting in part filed by Circuit Judge GEORGE CLIFTON EDWARDS, Jr.
Dissenting opinion filed by Circuit Judge WILKEY.
PER CURIAM:Judge Bazelon’s opinion constitutes the opinion of the court. Judge George Clifton Edwards, Jr., concurs in all but Part IV-D of the opinion, and Judge Wilkey concurs in only Part IV-D.
BAZELON, Senior Circuit Judge:These consolidated cases involve the continuing efforts of the Nuclear Regulatory *13Commission (NRC)1 to establish a system by which to consider and disclose the environmental impact of the uranium fuel cycle in compliance with the National Environmental Policy Act (NEPA).2 The present controversy centers upon the radiological effluents associated with the “back end” of the fuel cycle: the reprocessing,3 storage, and “disposal”4 of spent fuel and other wastes. At issue are three versions — the original,5 interim,6 and final versions7 — of the “Table S-3 Rule,” which provide a set of numerical values intended to reflect the environmental effects of the uranium fuel cycle.8 Under the Rule, Table S-3 is to be *16included in the environmental impact statement (EIS)9 of each proposed light water nuclear power reactor, and thereby substitute for repeated individualized consideration of the environmental impact of the fuel-cycle activities needed to support each plant.10
The issues in this case largely concern the use of numerical values to depict the environmental effects of fuel-cycle activities. By describing such effects in this manner, the issue arises whether there is more to the fuel cycle’s environmental impact than the bare numbers in the Table reveal, and, if so, whether licensing boards are prevented from looking beyond the Table to consider additional elements of the fuel cycle’s environmental impact. One omission from the Table is explicit recognition of the uncertainties that underlie the projected effluent releases. This omission is particularly glaring in the Table’s treatment of the long-term effects of solid high-level and transuranic wastes, which remain toxic for at least 250,000 years.11 The Commission ex*17pects to dispose of those wastes, perhaps by first reprocessing a portion of them, and in any event, by burying them in salt mines beneath the continental United States. The Tables indicate that the wastes will have no effect on the environment after they are sealed in salt mines.12 In effect, therefore, the Table S-3 Rules instruct all licensing boards, when analyzing the environmental impact of a particular plant, to conclusively assume that such wastes will emit no radiological effluents into the environment after final burial.
The second omission from the Table is a description of the health, socioeconomic, and cumulative effects of the projected releases. The Table describes effluent releases in units of radioactivity per year. It does not evaluate the actual effects of those releases on human health, or on social and economic well-being, as an increasing number of plants continue to produce wastes. The issue raised by this aspect of the Table is whether the Table S-3 Rule allows licensing boards to take evidence on and consider those effects in individual licensing proceedings.13 Finally, the effluent releases predicted in the Table are based on assumptions concerning the future availability of certain types of waste-management and disposal technology. The issue arises, therefore, whether the Commission has correctly found that the predicted releases represent technological goals that are economically feasible.
We conclude that the Table S-3 Rules are invalid because they fail to allow for proper consideration of the uncertainties concerning the long-term isolation of high-level and transuranic wastes, and because they fail to allow for proper consideration of the health, socioeconomic and cumulative effects of fuel-cycle activities. Therefore, we remand.
I. BACKGROUND
Although a great deal remains to be learned about radioactive wastes, this much is known: 1) many of the wastes remain extremely toxic for a very long time;14 *182) the NRC has yet to settle upon a method for permanently disposing of radioactive wastes;15 and 3) none of the current proposals for disposal is certain to succeed.16 For more than a decade, and in several different arenas, the Natural Resources Defense Council (NRDC) has sought to force the Commission to factor this knowledge into its licensing decisions, each of which represents a decision to create additional nuclear wastes.17
A. The Original Table S-3 Rule
In April 1971, the NRDC attempted to raise the environmental impact of nuclear waste as a relevant consideration in the operating license proceeding for the Vermont Yankee Nuclear Power Station.18 NRDC argued that such a licensing proceeding was the only decisionmaking stage at which the issue could be raised effectively, for at any later stage a commitment to produce potentially dangerous waste would have already been made. The Atomic Safety and Licensing Board (ASLB), however, rejected the NRDC’s request.19 It refused to admit evidence or permit questions on the environmental impact of Vermont Yankee’s spent fuel or other radioactive waste products.20
In June 1972 the Atomic Safety and Licensing Appeal Board (ASLAB) upheld the ASLB’s decision. The Appeal Board based its ruling on the fact that fuel-cycle effects were both speculative and remote. Because spent fuel and wastes would be removed from the plant, the Board stated, their effects should be considered in proceedings to license future reprocessing plants and waste *19repositories — not in Vermont Yankee’s licensing proceeding.21
In November 1972, the Commission responded to the ASLAB’s Vermont Yankee decision by publishing a notice of proposed rulemaking to determine whether, and if so, how, it should consider the environmental impact of the fuel cycle as it continued licensing nuclear facilities.22 The notice proposed two alternatives. The first alternative would have precluded any consideration of the environmental effects of radioactive wastes on the ground that those effects would not significantly affect the outcome of any licensing decision.23 The second alternative would have limited disclosure and consideration of the environmental effects of these wastes to a table of predetermined numerical values, which was reproduced in the notice.24 The values were standardized to represent the expected contribution of one 1,000 megawatt light water nuclear reactor. The Commission staff had already derived the necessary values and reported them in Table S-3 of the Commission’s Environmental Survey of the Nuclear Fuel Cycle.25 The Table contained entries for the environmental effects of most stages of the uranium fuel cycle, but it contained no entry for the radiological effluents from solidified high-level and transuranic wastes. These wastes, which must be isolated from the environment for hundreds of thousands of years, were expected to be buried eventually in a federal repository somewhere beneath the continental United States. The original Table contained no values for the environmental impact of this final stage of the fuel cycle because the Commission staff believed that technology would be developed by which to isolate the wastes from the environment for an indefinite period of time.26 Thus, both alternative proposals were based on the assumption that long-term waste-disposal systems, yet to be developed, would work perfectly. On the basis of such assumptions, the Commission proposed to mandate that licensing boards presume, in their decisions to grant nuclear power plant licenses, that there would be no risk of waste-disposal failure.
On April 16, 1974, the Commission adopted the Table S-3 alternative and amended its NEPA regulations to allow for consideration of the environmental effects of the uranium fuel cycle by including Table S-3 in the EIS for each light water nuclear reactor.27 The Rule also provided that the environmental effects of the fuel cycle “shall be as set forth in Table S-3,” and that “no further discussion of such environmental effects shall be required.”28 In the meantime, the Vermont Yankee plant itself *20had been granted a full-term, full-power operating license,29 and the Appeal Board had expressly declined to reexamine their earlier decisions precluding inquiry into the environmental impact of the fuel cycle.30
The NRDC challenged both the Table S-3 Rule and Vermont Yankee’s license in this court.31 We found that the procedures that the Commission employed in adopting the Rule failed to generate a record sufficient to support the optimistic conclusion that high-level and transuranic wastes would be isolated from the environment for the requisite number of centuries.32 The court noted that the portion of the rule precluding inquiry into these matters was based on a few conclusory reassurances from a member of the Commission’s staff who was not questioned or cross-examined.33 That part of the Rule was therefore held to be arbitrary and capricious, and was set aside and remanded. We also held that in the absence of a valid generic rule, the environmental impact of fuel-cycle activities must be considered in individual licensing proceedings. Because the Commission had granted Vermont Yankee’s operating license without such consideration, we remanded the order granting that license to await the outcome of further generic proceedings.34 These decisions were rendered July 21, 1976.
B. The Supreme Court’s Decision
In April 1978, the Supreme Court reversed this court’s decision to set aside the original Table S-3 Rule,35 finding that we had ■ overturned the Rule because of deficiencies in the Commission’s rulemaking procedures.36 The Court held that if an agency complies with the procedures required by statute, a rule may be struck down because of procedural shortcomings only in unusual circumstances.37 The Supreme Court agreed, however, that the Rule should be vacated if it lacks support in the administrative record, and remanded the case to us “so that the Court of Appeals may review the rule as the Administrative Procedure Act provides.”38 That case, No. 74-1586, is, therefore, among the four consolidated cases now before us.
The Supreme Court also noted that the Commission had promulgated an interim Rule pending the issuance of a final Rule. It stated that this court, on remand, could consolidate the challenge to the original *21Rule with the appeal from the interim rule-making proceeding, which was already pending. The Court further noted that this court could decide those cases on the basis of an expanded record, which is what we have done.39
C. The Interim Table S-8 Rule
On October 18, 1976, in response to this court’s decision but prior to the Supreme Court’s remand, the Commission initiated proceedings to review the waste-management and disposal aspects of the Table S-3 Rule, and to develop a revised and adequately supported fuel-cycle rule.40 The Commission announced that a special Task Force had completed a revised Environmental Survey of the Reprocessing and Waste Management Portions of the LWR Fuel Cycle,41 and invited comments on the Survey and the revised Table S-3, both of which would constitute the basis of a final or interim rule.42 Also in response to this court’s ruling, the Commission had reconvened a licensing board to consider suspending the operating license of Vermont Yankee, which, by then, was fully operational.43 The Commission suspended that proceeding, however, when it decided that an interim rule, similar to the original, could issue shortly.44
In announcing that it was reconsidering the Table S-3 Rule, the Commission stated:
[T]here are still uncertainties in areas such as the effect of waste presence on repository stability; the probabilities and consequences of various types of intrusive acts by humans; the availability of data to be used in modeling studies; the design and regulatory actions needed to minimize possibilities of repository failure; projection of future societal habits and demography; and, finally, the relative importance of the various potential initiating events. Research programs are underway which should resolve most of these uncertainties over the next few years.45
Nonetheless, the Survey and proposed Table S-3 still provided that solidified high-level and transuranic wastes, which were to be buried in a permanent repository, would have no effect on the environment. Unlike the original Table, the revised Table S-3 did contain an entry for solidified high-level and transuranic wastes. It provided that 11 million curies of radiation would be released from the solidified waste per reference reactor year. Under the “maximum effect” column, however, the Table simply stated “Buried at Federal Repository.”46 The materials referenced by the footnotes to the Table make clear that this meant that the radiation would remain wholly contained within the repository once the repository is sealed.47
The NRDC filed comments stating that the Survey and Table were based on only an assumption that technology could be developed to isolate long-term wastes from the environment, and that the consequences of the failure of such development were left unanalyzed. The NRDC’s comments also stated that the numerical values in the Table failed to reveal the health effects of environmental impacts or the cumulative *22effects of the continuous creation of radioactive wastes.48 In addition, the State of New York filed comments objecting to the NRC’s failure to consider the economic feasibility of the projected waste-management and disposal methods.49 The Commission staff and, ultimately the Commission, rejected all of these complaints and refused to modify the proposed rule.50
In March 1977, on the basis of the revised Survey and the comments received, the Commission promulgated an interim Table S-3 Rule,51 which, as indicated in the Commission’s notice, explicitly stated that releases from the solidified high-level and transuranic wastes would remain buried-in a repository and, therefore, have no effect on the environment.52 Like the original Table S-3 Rule, the interim Rule stated that the environmental effects of the fuel cycle “shall be as set forth in Table S-3,” and that “[n]o further discussion of such environmental effects shall be required.”53 The Commission amended the interim Rule in April 1978, however, to allow licensing boards to consider environmental impacts not specifically “addressed by the Table.”54
In April 1977, the Commission issued an order directing the Appeal Board to reconsider the cost-benefit balances which had been struck in the licensing proceedings of Vermont Yankee and several other plants.55 The following July, the Board ruled that the values in Table S-3 did not change the cost-benefit balance of the Vermont Yankee plant, particularly because the plant was already operating.56 In connection with this reconsideration, the Board expressed *23concern over the proper interpretation of the Table S-3 entry for high-level and transuranic wastes, given the fact that the Commission openly admitted that the assumption of no environmental effect was uncertain.57 The Board strongly implied that its interpretation of that aspect of the Table could alter the outcome of particular decisions.58 Ultimately, however, the Appeal Board concluded that the Commission had intended it to assume that there would be no effluent releases from the permanent repository, once it is sealed.59
In No. 77-1448, the second of the four cases now before this court, NRDC challenges the interim Rule.60
D. The Final Table S-8 Rule
In May 1977, following the adoption of the interim Rule, the Commission reopened hearings to determine whether the interim Rule should be made final or altered in any way.61 The hearings employed some procedures not used in the original rulemaking, and participants were given an additional opportunity to comment on the Hearing Board’s recommendations to the Commission.62 The NRDC criticized the Rule for the failure of the numerical values to communicate either the uncertainties of underlying assumptions, or the health, cumulative, or socioeconomic effects of waste-management and disposal activities.63
The States of New York and Wisconsin again objected to the Commission’s failure to analyze the economic viability of the proposed waste-management and disposal methods. The Commission staff initially argued that the economics of the proposed *24methods should not be considered.64 The Hearing Board, however, and later the Commission, ordered that evidence concerning the economic feasibility of the proposed technologies be considered in the rulemaking.65 The States of New York, Ohio, and Wisconsin submitted such evidence, criticizing the cost projections and the economic feasibility of the anticipated disposal methods.66 In response, the Commission staff and industry groups submitted evidence in support of the cost estimates used.67
In July 1979, the Commission promulgated the final Table S-3 Rule.68 The final Rule made only minor adjustments in the Table’s numerical values and left unchanged the assumption that radiological effluents from solidified high-level and transuranic wastes would have no effect on the environment once sealed in a federal repository.69 The Commission also stated its intention to add an explanatory narrative to Table S-3 to convey the significance of the numerical values.70 In addition, the Commission confirmed its intention that the Table should be supplemented by individual presentations on the health, socioeconomic, and cumulative effects of fuel-cycle activities, pending the adoption of such a narrative.71
As it had done before, the Commission noted the uncertainties regarding both the likelihood of finding a site for a permanent repository and the likelihood that the repository will perform as expected.72 Nonetheless, the Commission explicitly rejected the option of expressing uncertainties in the Table S-3 Rule. The Commission stated:
In view of the uncertainties noted regarding waste disposal, the question then arises whether these uncertainties can or should be reflected explicitly in the fuel cycle rule. The Commission has concluded that the rule should not be so modified. On the individual reactor licensing level, where the proceedings deal with fuel cycle issues only peripherally, the Commission sees no advantage in having licensing boards repeatedly weigh for themselves the effect of uncertainties on the selection of fuel cycle impacts for use in cost-benefit balancing. This is a generic question properly dealt with in this rulemaking as part of choosing what impact values should go into the fuel cycle rule. The Commission concludes, having noted that uncertainties exist, that for the limited purpose of the fuel cycle rule it is reasonable to base impacts on the assumption which the Commission be*25lieves the probabilities favor, i.e., that bedded-salt repository sites can be found which will provide effective isolation of radioactive waste from the biosphere.73
Thus the Commission’s final Rule, like its two predecessors, does not permit licensing boards to consider either the risk that permanent waste management facilities will not be developed or the risk that they will fail to perform as intended if they are developed. In addition, the Commission accepted the Hearing Board’s recommendation and conclusion that the projected facilities were economically feasible because they were not “outlandishly expensive.”74
In 79-2110, the State of New York challenges the final Rule on the ground that the Commission’s finding of economic feasibility is arbitrary and capricious. In 79-2131, the' NRDC challenges the final Rule because it inadequately discloses, and fails to allow proper consideration of, the uncertainties underlying the environmental impacts embodied in the final Table.
II. STANDARDS OF REVIEW
As the Supreme Court has stated, we must review the Table S-3 Rule under the standards provided in the Administrative Procedure Act (APA).75 The applicable provision of the APA is section 706(2)(A), which provides that a reviewing court shall “hold unlawful and set aside agency action, findings and conclusions found to be — arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.”76 In this case, both the “arbitrary and capricious” and the “not in accordance with law” standards are appropriate. Under the latter standard, we must determine whether the Table S-3 Rule, as applied to individual licensing decisions, violates NEPA,77 while under the “arbitrary and capricious” standard, our task is to “consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.”78 Of course, other laws must give content to the phrase “relevant factors.” Because the Table S-3 Rule governs the preparation of environmental impact statements for individual nuclear reactors, NEPA provides the relevant factors that the Commission was required to consider in formulating the Rule. Therefore, there is an area of overlap between the “arbitrary and capricious” standard and the “in accordance with law” standard.
III. NEPA
It is well settled that the licensure of a nuclear power plant constitutes a “major Federal action[] significantly affecting the quality of the human environment.”79 Therefore, there can be no doubt that the Commission must comply with section 102(2)(C) of NEPA80 prior to licensing such a plant. That section provides:
*26The Congress authorizes and directs that, to the fullest extent possible ... all agencies of the Federal Government shall . . . include in every recommendation or report on proposals for legislation and other major Federal actions significantly affecting the quality of the human environment, a detailed statement by the responsible official on—
(i) the environmental impact of the proposed action, [and]
(ii) any adverse environmental effects which cannot be avoided should the proposal be implemented, [and]
* * * * * *
(v) any irreversible and irretrievable commitments of resources which would be involved in the proposed action should it be implemented.
Furthermore, as the Supreme Court has stated,
[i]t is hard to argue that [nuclear wastes] do not constitute “adverse environmental effects which cannot be avoided should the proposal be implemented,” or that by operating nuclear power plants we are not making “irreversible and irretrievable commitments of resources.”81
Therefore, the Commission must fully consider and disclose the environmental impact of nuclear wastes during the course of its decisions to license nuclear power plants.82
It is well recognized that section 102(2)(C) of NEPA has a dual function: Its first function is “to inject environmental considerations into [federal agencies’] decisionmaking processes];”83 and its second function is to inform the public of the nature of the agencies’ decisions, and to “inform the public that [agencies have] considered environmental concerns in [their] decisionmaking processes].”84 With respect to these statutory purposes, the Supreme Court has stated, and recently reiterated 85 that “[t]he thrust of § 102(2)(C) is . . . that environmental concerns be integrated into the very process of agency decisionmaking. The ‘detailed statement’ it requires is the outward sign that the environmental values and consequences have been considered during the planning stage of agency actions.”86 In Calvert Cliffs’ Coordinating Committee, Inc. v. AEC,87 one of the first cases decided under NEPA, this court analyzed these aspects of the Act in further detail. We stated that “Compliance to the ‘fullest’ extent possible would seem to demand that environmental issues be considered at every important stage in the decisionmaking process concerning a particular action — at every stage where an overall balancing of environmental and nonenvironmental factors is appropriate and where alterations might be made in the proposed action to minimize environmental costs.”88 Thus, decisionmaking under NEPA must be organized in such a manner that all of the reasonably foreseeable environmental effects of a proposed action enter into an agency’s decision to take the action.89 The environmental impact statement, the centerpiece of NEPA’s procedural requirements, both facilitates and provides a record of this decisionmaking process. As we noted in Calvert Cliffs, two related pur*27poses of the EIS are to provide evidence that the mandated decisionmaking process has taken place, and to “allow [] those removed from the initial process to evaluate and balance the [environmental] factors on their own.”90 The decisionmaking structure mandated by section 102(2)(C) is not inherently flexible, nor are the particular requirements that it entails discretionary.91 It is the role of the courts to ensure that agencies comply with these provisions of NEPA.92
The NRC’s promulgation of the Table S-3 Rule represents both an attempt to implement, and a limitation on other means of implementing, these aspects of NEPA.93 The validity of the Rule, therefore, turns largely on whether the decisionmaking structure it mandates conforms to the procedures, and thereby poses no obstacles to the substantive calculus, required by NEPA.94 Specifically, we must determine whether the Table S-3 Rule (1) violates NEPA; or (2) is arbitrary and capricious, given the considerations made relevant by NEPA.
IV. ANALYSIS
Petitioners, the NRDC and the State of New York, challenge the Table S-3 Rules on three grounds. First, the NRDC argues that the original, interim and final versions of the Table S-3 Rule are arbitrary and capricious and not in accordance with NEPA because they preclude proper consideration and disclosure of uncertainties that underlie the Tables’ numerical values. Second, the NRDC argues that the original and interim Table S-3 Rules are arbitrary and capricious and not in accordance with NEPA because they preclude proper consideration and disclosure of the tangible environmental effects of the waste created by nuclear power plants. Finally, the State of New York argues that the release levels projected in the final Table S-3 Rule are based on a finding of economic feasibility that was arbitrary and capricious.
We conclude that the Table S-3 Rules are arbitrary and in violation of NEPA because they fail to allow for consideration of uncertainties underlying the assumption that no radiological effluents will be released into the biosphere once wastes are sealed in a permanent repository. Similarly, we conclude that the original Rule and the interim Rule, prior to amendment, are arbitrary and in violation of NEPA in their failure to allow consideration of health, socioeconomic, and cumulative ef*28fects of fuel-cycle activities. We also conclude, however, that the Commission’s finding that the predicted waste-management and disposal methods would be economically feasible was neither arbitrary nor capricious.
A. The Zero-Release Assumption and Uncertainty Over Permanent Storage
The original, interim, and final Rules all prohibit licensing boards from considering the uncertainties surrounding the Table S-3 Rules.95 In particular, no challenges were allowed to the assumption that a suitable final repository site would be found, constructed, and would operate flawlessly.96 The assumption underlying Table S-3 that no radiological effluents will be released from the permanent repository, once the repository is sealed, can be interpreted in either of two ways: It can be read as a factual finding, or, as a decisionmaking device by which the Commission retains exclusive responsibility for considering the uncertainties concerning long-term waste disposal. Particularly in the interim and final Rules, it appears that the latter is the correct characterization.97 Nonetheless, for the sake of completeness, we review the validity of the zero-release assumption under each interpretation.98
1. The Zero-Release Assumption as a Factual Finding
As stated in Part II, above, a reviewing court must set aside an agency finding as arbitrary if it determines that the finding is not based upon consideration of relevant factors or if it is based on a clear error of judgment.99 The “relevant factors” in this case are those whose consideration is mandated by NEPA.
Among the environmental costs that an agency must consider under section 102(2)(C) are significant environmental risks — probabilities or possibilities of environmental damage.100 Such risks may be *29present due to the underlying randomness of nature. Or they may be due to human uncertainty over either the character of both random and nonrandom phenomena or the ability of future technology to cope with those phenomena. Regardless of the source, environmental risks are environmental costs that must be factored into the NEPA calculus.101 They must also be included in an EIS. Where such risks are attributable to a lack of knowledge, the agency can be asked to do no more than to reveal that which it knows and that which it does not know. It may not be permitted, however, to do any less.102
An agency could state in an EIS, as a matter of factual prediction, that a particular environmental effect will not have to be endured as a result of a proposed action. Of course, if it believes that to be the case, it can omit entirely any discussion of the would-be effect. Similarly, an agency could provide, by generic rule, that a particular environmental effect will not be caused by a class of actions, and that the effect should not be addressed in individual environmental impact statements. Because the risk of an environmental effect is the overriding “relevant factor” that an agency must consider in making such a factual finding, the agency may treat an environmental effect in this manner only if it finds that there is no significant risk-that the environmental effect will occur. It may not do so if it finds only that the effect is unlikely to occur. Otherwise, NEPA’s requirement that agencies consider and disclose uncertainty would be subverted. When faced with uncertainty concerning an environmental effect, an agency could evade its obligations under NEPA simply by finding *30that the effect will not occur. Thus, in the context of compliance with NEPA’s procedural requirements, a court must interpret an agency’s finding of no environmental effect as a finding that there is no significant risk of an environmental effect.
It is against that background that we review the Commission’s finding that nuclear wastes that are sealed in a permanent repository will have no impact on the environment. When read as a finding that such wastes pose no significant risk of environmental damage, we conclude that the finding represents a “clear error in judgment.” 103
In each version of the Table S-3 Rule, the Commission based its zero-release assumption on a prediction that technology would be developed by which to isolate long-lived wastes from the biosphere indefinitely. In the Environmental Survey,104 which constituted the technical basis of the original Table S-3, the Commission staff reported that “[i]t was planned to construct a Federal repository in a salt mine for long-term geological storage of solid high-level wastes by the mid-1970’s. However, subsequent events have deferred the site selection and construction of such a facility.” 105 The plan at that time, therefore, was to build a Retrievable Surface Storage Facility (RSSF) in which to store the wastes until a permanent repository became available. The Environmental Survey said no more than that the permanent repository that would be developed “is intended to isolate the waste from man and the biosphere.” 106 A few years later, in the report that formed the basis of the interim and final Rules, the Commission staff continued to assume that a bedded-salt repository could be developed, but noted that other systems were also under consideration.107 The other systems included the use of alternative geological media for isolation, and the “eliminat[ion of] portions of the wastes from existence on earth.”108 The Commission still had not worked out the technological details of any permanent disposal techniques. Nonetheless, it concluded that a system would be developed that would fully protect the environment. Until such a system is developed, the Commission still plans to store waste in an RSSF.109
Scientists with the United States Geological Survey testified to the technical uncertainties surrounding the zero-release assumption.110 Similarly, the Report to the President by the Interagency Review Group on Nuclear Waste Management111 pointed out that risk assessments “based on idealized repository characteristics . . . are subject to significant uncertainties,”112 and concluded that the “zero release of radionuclides [from a permanent repository] cannot be assured.”113 In addition, the record indicates that serious concerns were raised over the likelihood of developing the human institutions or political consensus necessary to establish and maintain the hypothesized fa*31cilities.114 The IRG Report noted, in fact, “that the resolution of institutional issues may well be more difficult than finding solutions to remaining technical problems.” 115 Moreover, even the Commission’s own staff recognized that there were many uncertainties surrounding the assumption that a permanent repository could be developed that would fully contain the effluents emitted from high-level and transuranic wastes.116
There is no need, however, to comb the record further for evidence of uncertainty, for the Commission itself has explicitly — although somewhat belatedly — acknowledged its presence. In its Statement of Consideration for the final Rule, “the Commission note[d] and agree[d] ... that areas of uncertainty remain regarding both the likelihood of finding a site and the probability that it will perform as expected.” 117 The Commission nevertheless accepted its staff’s zero-release assumption, stating that the staff’s conclusion that there would be no release from the sealed repository had a “reasonable basis.” Moreover, revealing even more of the uncertain nature of its judgment, the Commission concluded that the evidence, although “tentative” and general in nature, “favors the view that suitable sites can be found.”118
The evidence in the record and the Commission’s own recent statements indicate the existence of such uncertainty concerning permanent disposal of high-level and transuranic wastes that the zero-release assumption, taken as a finding of fact, cannot stand. If read as a finding of no significant risk, which we acknowledge may not have been its intent, the zero-release assumption represents a self-evident error in judgment. Therefore, we conclude that the finding is arbitrary and capricious.
2. The Zero-Release Assumption as a Decisionmaking Device
As stated above, the Commission’s zero-release assumption is probably characterized better as a decisionmaking device than as a finding of fact, particularly in the interim and final Rules. By instructing licensing boards to assume that nuclear waste will have no impact on the environment once it is sealed in a repository, the Table S-3 Rule has served to allocate to the Commission sole responsibility for considering the risk that long-lived wastes will not be disposed of with complete success. Under that decisionmaking scheme, the Commission considered the possibility of unsuccessful isolation of wastes prior to promulgating the Table S-3 Rule. The Rule itself represents the Commission’s statement that *32it has done so, and that licensing boards should go forward licensing plants without duplicating its effort.
Just as the Commission avoided describing the zero-release assumption as a finding of fact, it did not describe the assumption in precisely these terms either. In more recent years, however, the Commission has, at various times, seemed to indicate that this is essentially its view of the assumption. For instance, in its Statement of Consideration for the Final Rule, the Commission stated:
In view of the uncertainties noted regarding waste disposal, the question then arises whether these uncertainties can or should be reflected explicitly in the fuel cycle rule. The Commission has concluded that the rule should not be so modified. On the individual reactor licensing level, where the proceedings deal with fuel cycle issues only peripherally, the Commission sees no advantage in having licensing boards repeatedly weigh for themselves the effect of uncertainties on the selection of fuel cycle impacts for use in cost-benefit balancing. This is a generic question properly dealt with in this rulemaking as part of choosing what impact values should go into the fuel cycle rule. The Commission concludes, having noted that uncertainties exist, that for the limited purpose of the fuel cycle rule it is reasonable to base impacts on the assumption which the Commission believes the probabilities favor, i.e., that bedded-salt repository sites can be found which will provide effective isolation of radioactive waste from the biosphere.119
Our inquiry, therefore, must focus on whether the decisionmaking device of the Rule — including the Commission’s investigation, analysis, and deliberation prior to the Rule’s promulgation — has provided the type of consideration and disclosure of uncertainties that NEPA requires.
In general, an agency in the position of the Commission is free to implement NEPA through generic rulemaking. If certain types of environmental costs are common to a class of actions, NEPA does not require that an agency engage in duplicative and possibly inconsistent individual determinations, but allows it, in the alternative, to conduct a single rulemaking to determine generic values to be considered together with case-specific costs and benefits in individual proceedings.120 Similarly, if there are both costs and benefits common to a class of individual actions, an agency is free, not only to determine generic values for those costs and benefits, but also to weigh the costs and benefits against each other to produce a generic “net value.” To the extent that certain costs and benefits cannot be compared on a single scale, the generic determination of net value is necessarily more complex. It may involve instructing case-specific decisionmakers to insert fictional values into their cost-benefit analyses. A generic rule might be promulgated, for instance, designating that certain entries on one side of an individual actions’ ledger should be treated as zero, or nonexistent, and assigning offsetting values to certain entries on the other side of the ledger. The agency may even determine that certain classes of generic costs and benefits so balance each other that both can be treated as zero for purposes of individual decisions. In the abstract, such generic structuring of individual decisionmaking is acceptable as long as it is based on the agency’s reasoned judgments about how the generic costs and benefits weigh against each other.121 In the course of such a generic rulemaking, however, the agency must consider and disclose the actual environmental effects it has assessed in a manner that will ensure that the overall process, including both the generic rulemaking *33and the individual proceedings, brings those effects to bear on decisions to take particular actions that significantly affect the environment. If the overall decisionmaking process allows for this type of consideration — which is precisely that described in Calvert Cliffs’122 — the agency may then direct case-specific decisionmakers to abide by the generic rule, and thereby preclude their reevaluation of either the underlying determinations of costs and benefits or the process by which those costs and benefits were assessed against each other.
Under the Table S-3 Rule, licensing boards consider virtually all factors relevant to the construction and operation of a nuclear power plant. The zero-release assumption, however, excludes from their consideration two factors: 1) uncertainty concerning the integrity of the permanent repository, if such a repository is ever built; and 2) uncertainty over whether and when such a repository, or equivalent system of disposal, will be developed. These uncertainties reflect two environmental costs of licensing a plant. The first cost is the risk that wastes created by the plant will eventually damage the environment by emitting radiological effluents from a faulty permanent repository. The second cost is the risk that waste created by the plant will have to remain in another type of repository — possibly on site123 — and emit radiological effluents prior to permanent disposal, if such disposal ever comes about. Hence, the zero-release assumption prevents licensing boards from considering two environmental costs.
The Commission argues that its consideration and disclosure of these costs prior to promulgating the Rule was sufficient under NEPA, and that the licensing boards need not reconsider them. We disagree. Although the Commission did consider these uncertainties, it did not do so in a manner that would allow licensing decisions to be affected — either directly or indirectly — by the risk that nuclear waste will not be successfully isolated from the environment indefinitely. After recognizing that there are uncertainties concerning the permanent storage of nuclear wastes, the Commission simply ruled that licensing decisions should be made on the basis of cost-benefit analyses that omitted the costs represented by those uncertainties. It did not rule that the costs were insignificant,124 nor did it rule that they were outweighed by generic benefits that would also be excluded from licensing boards’ consideration. In effect, therefore, the Commission directly contravened NEPA’s requirement that environmental costs be considered “at every stage where an overall balancing of environmental and nonenvironmental factors is appropriate.” 125 The risks entailed by the possible failure to develop a successful waste-dispos*34al system were never part of any “balancing.” They were considered alone, in a vacuum, and then excluded from the licensing boards’ balancing. The process that began with the proposal of the Table S-3 Rule does not allow the uncertainties concerning permanent storage to play a role in the ultimate licensing decision. That omission, and hence, the Rule, which causes it, constitutes a blatant violation of NEPA.
In arguing that it adequately considered the uncertainties surrounding long-term storage, the Commission emphasizes the fact that these uncertainties represent generic costs.126 The Commission recognizes that cost-benefit balances differ among individual plants and that these uncertainties are relevant costs, particularly if a plant’s cost-benefit balance is close.127 Nonetheless, the Commission asserts that “[it remains] a generic question how to weigh these uncertainties [in individual cases].”128 To the extent that the Commission is arguing that the uncertainties can be assessed generically, and that the attendant risks can be measured generically, we agree. As we have stated above, however, the Commission cannot find the environmental cost represented by the uncertainties to be zero unless their cost is, in fact, zero.129 Similarly, although the Commission can assess generic costs in a generic rulemaking,' it must, in some manner, factor its assessment into ultimate decisions to license plants.130 One way in which the Commission could do so would be to follow the course that it has taken with respect to other environmental costs of licensing nuclear power plants: It could assess and evaluate the uncertainties and attendant risks, and instruct licensing boards to consider and disclose them in a uniform manner. To the extent that the Commission is arguing, however, that generic costs need not be considered at all within the context of balancing the costs and benefits of licensing decisions, we strongly disagree. The fact that certain environmental costs can be assessed generically does not imply that they can be excluded entirely from the balancing process that must precede an agency’s decision to take a major action.131
The Commission also argues that its disclosure of uncertainties in its Statements of Consideration and in its staff reports, satisfies NEPA’s requirement of a “detailed statement.” We recognize that the NRC, particularly in promulgating the final Rule, has disclosed the nature of many of the uncertainties surrounding the storage of long-lived wastes.132 NEPA, however, requires an agency to do more than to scatter its evaluation of environmental damage among various public documents. As stated above, an agency must disclose environmental costs — including uncertainties concerning such costs — in a manner that proves to the public that the agency has properly considered the environmental costs of its action.133 In this case, because the Commission has not properly considered such costs, the issue whether it has disclosed enough raw information is beside the point.
In sum, when viewed as an administrative decisionmaking device, the zero-release assumption of Table S-3 constitutes a violation of NEPA. Under section 706(2)(A) of *35the APA, it is, therefore, “not in accordance with law.”134
Alternatively, the same result can be reached under the arbitrary and capricious standard.135 Under that standard, as stated above, our inquiry focuses upon whether the zero-release assumption is based on “consideration of the relevant factors.”136 Under NEPA, significant uncertainty surrounding the environmental effect of a proposed action is relevant to an agency’s decision to rule generically that the effect will not occur. For an agency to go forward in the face of significant uncertainty and issue such a rule indicates either a failure to consider a relevant factor or a clear error in judgment. Because that is precisely what the Commission did in promulgating the Table S-3 Rule, we could also conclude that its action was arbitrary and capricious.137
B. Consideration of Other Uncertainties
The NRDC seems to argue that Table S-3 masks uncertainties underlying waste-management activities other than permanent disposal.138 It has not, however, indicated where those uncertainties lie or what values in the Table obscure them. Similarly, the Commission has responded to the NRDC by focusing primarily on the uncertainties surrounding the successful development of a permanent repository.139 As a result, we face considerable difficulty reviewing this issue.
Table S-3 lists several gaseous and liquid radiological effluents along with low-level, high-level and transuranic solid wastes.140 The gaseous and liquid effluent releases are expected to occur at the early stages of the waste-management process, before the wastes are solidified and prepared for per*36manent disposal. The staff reports supporting the Table S-3 Rules indicate that there is indeed uncertainty concerning the predictions of these releases. In contrast to its method of projecting releases from the solid waste stored in a permanent repository, however, it appears that the Commission used “worst case” estimates in determining Table S-3’s values for gaseous and liquid releases. The Commission assumed, for example, that all of the tritium, krypton-85 and carbon-14 in spent fuel would be discharged into the environment during reprocessing.141 The Commission also seems to have used equally conservative assumptions in determining Table S-3’s values for releases of iodine-129, iodine-131 and other radioactive gases.142 One issue before us, therefore, is whether the Commission violated NEPA in treating these uncertainties in this manner, rather than by representing uncertainty itself in the Table. We hold that it did not.
As we have emphasized above, NEPA requires an agency to consider the environmental risks of a proposed action in a manner that allows the existence of such risks to influence the agency’s decision to take the action.143 An agency can do this by having the appropriate decisionmakers consider all that is known and unknown about the risks before deciding whether to take an action. Or, it can organize its decisionmaking process in such a manner that the appropriate decisionmakers consider only the upper bound of reasonably foreseeable environmental costs. If, after considering that level of environmental damage, the decisionmakers conclude that the proposed action is worth its societal costs, full account will have been taken of the action’s environmental impact. Similarly, if the upper bounds of environmental risks are disclosed in an EIS, Congress, the public, and any interested agency can effectively assess for themselves whether the agency has proposed an action that is not worth its environmental costs. Either method of considering and disclosing uncertainties surrounding an environmental effect is acceptable under NEPA.144 Thus, to the extent that the Commission has listed the upper bound of reasonably foreseeable gaseous and liquid effluent releases in Table S-3, we hold that it has complied with NEPA.
It is impossible for this court to determine whether there are other numerical values in Table S-3 that mask uncertainties without using such conservative heuristic devices as a worst-case scenario. We expect, however, that the Commission’s consideration of uncertainties underlying the zero-release assumption will encompass any significant uncertainties underlying other values in the Table.145
C. Consideration of Health, Socioeconomic and Cumulative Effects
Petitioner NRDC also challenges the original and interim Rules for failing to allow proper consideration or disclosure of the actual environmental impact of the fuelcycle.146 In all versions of the Rule, Table S-3 lists the environmental effects of the fuel cycle in terms of the quantity of land, water, and energy used, and of heat, chemicals and radioactivity released. It does not reveal the meaning of those impacts in terms of human health or other environmental values. The Table does not, for instance, indicate the number of cancer deaths or genetic defects to be expected. Nor does it refer to potential social, psychological, and economic disruptions that might accompany the siting of waste repositories, the construction of reprocessing and dispos*37al facilities, or the maintenance of safety and security over long periods of time. Furthermore, the Table, which is based on the incremental impact of one prototypical reactor operating for one year, assumes that there will be no cumulative effects of licensing scores of reactors each of which will remain in operation for thirty to forty years.
There can be no dispute that NEPA requires the health, socioeconomic and cumulative impacts of a proposed action to be disclosed in an EIS.147 Otherwise, the statement could not fulfill its purpose of informing the public, Congress, and other decisionmakers of the environmental impact of the action.148 Nor could the environmental cost-effectiveness of a proposed action be compared to that of alternative actions if the environmental effects of each are not disclosed in such commensurable terms. In this case, therefore, it is not releases of curies that Congress wanted disclosed; it is the effects, or environmental significance, of those releases. Those effects are defined by the CEQ regulations to include:
ecological (such as the effects on natural resources and on the components, structures, and functioning of affected ecosystems), aesthetic, historic, cultural, economic, social, or health, whether direct, indirect, or cumulative.149
Hence, the issue before us is whether the provisions of the original and interim Rules that governed the use of Table S-3 allowed for proper consideration of these environmental values.
Our resolution of that issue is complicated by the fact that the relevant provisions of the Rule have undergone several trans-. formations since their initial promulgation. The original and interim Rule provided that “the contribution of the environmental effects of . . . fuel cycle activities shall be as set forth in the following Table S-3 ... [and] no further discussion of such environmental effects shall be required.”150 Prior to 1977, individual licensing boards interpreted this clause to preclude their consid*38eration of the environmental significance of the numerical values listed.151 In its January 1977 Hartsville decision, however, the Commission’s Appeal Board ruled that a licensing board must consider the relative health effects of a proposed nuclear power plant and a coal-fired alternative.152 Then, in April 1978, the Commission promulgated a “clarifying” amendment to the interim Table S-3 Rule which specifically provided that health effects could be considered by individual licensing boards.153 In addition, the Commission broadened the scope of licensing boards’ consideration by amending the restriction on such consideration to read: “No further discussion of the environmental effects addressed by the Table shall be required.”154 The amended interim Rule did not, however, explicitly address the consideration of any type of impact other than health effects. Ultimately — or perhaps penultimately — in the final Rule, which the NRDC does not challenge, the Commission resolved the issue by requiring licensing boards to consider the socioeconomic and cumulative effects in addition to the health effects of the releases projected in the Table.155 At that point, the Commission also indicated that it would conduct further proceedings to determine such health, socioeconomic and cumulative effects on a generic basis, and thereby eliminate the need for case-by-case consideration.156
The NRDC argues that the original, the interim and the amended interim Rules violated NEPA by preventing licensing boards from considering the health, socioeconomic and cumulative effects of fuel-cycle activities.157 The Commission responds by arguing that each version of the Rule complied with NEPA because none explicitly precluded such consideration. We find the Commission’s argument unpersuasive.
The original Rule and the interim Rule, prior to its amendment, stated that “the contribution of the environmental effects of ... fuel cycle activities shall be as set forth in the following Table S-3 . . . [and] no further discussion of such environmental effects shall be required.”158 The term “such” refers to the phrase “the contribution of the environmental effects of . . . fuel cycle activities.” Therefore, the Rules unambiguously stated that a licensing board’s consideration and an EIS’s disclosure of the environmental effects of fuel-cycle activities could be limited to the information contained within the four corners of Table S-3. Moreover, it appears that the impact of the Rules was to preclude further consideration or disclosure. The background of the Rules makes this conclusion even clearer. Recall that the proposal for the original Rule suggested two ways of dealing with the environmental effects of fuel-cycle activities.159 One approach was to ignore those environmental effects in individual licensing proceedings as the Appeal Board had allowed in Vermont Yankee; and the other, which was adopted, was to factor those effects into individual cost-benefit analyses by using Table S-3. Since the Table S-3 Rule was adopted as an alternative to a rule that would have permitted no consideration or disclosure of the environmental effects of the fuel cycle, it was reasonable for licensing boards and interested parties to conclude that the Rule limited consideration of such environmental effects to the introduction of the Table.160 *39The Commission itself has recognized as much. For instance, in promulgating the final Rule, the Commission stated that the Table S-3 Rule “at least initially was apparently interpreted as cutting off further discussion of fuel cycle impacts” 161 and that
the rule in practice [was applied] as allowing fuel cycle impacts to be addressed in reactor licensing proceedings solely by the formal act of displaying Table S-3 in impact statements, with no further discussion. In particular, impact statements prepared by the staff did not analyze fuel cycle impacts in terms of health effects which might be caused by the radioactive releases tabulated in the rule and did not discuss socioeconomic or cumulative impacts.162
The original Rule and the interim Rule, prior to its amendment, thus effectively eliminated the consideration and disclosure of the health, socioeconomic and cumulative impacts of fuel-cycle activities. The NRC, therefore, in promulgating the original and interim Rules violated NEPA.163
The Commission argues that the Appeal Board’s Hartsville164 decision eliminated any preclusive effect that the Rules may have had on the consideration of health, socioeconomic and cumulative impacts.165 We disagree. First, the decision in Harts-ville was itself ambiguous. Although holding, in general, that a licensing board had to examine the comparative health effects of a nuclear plant and the alternative of a coal-fired plant, the Appeal Board specifically refrained from considering any environmental effect of waste-management and disposal activities, awaiting direction from the Commission in the then-pending interim Table S-3 Rule.166 Second, the Hartsville decision only addressed health effects in the context of comparing a nuclear plant with a coal-fired alternative.167 Finally, the Hartsville decision did not address cumulative or socioeconomic effects. Thus, even after Hartsville, it remained apparent that the consideration of health, and particularly, cumulative and socioeconomic impacts of the fuel cycle was precluded by the Table S-3 rules.168
*40Not until the Commission’s amendment of the interim Rule did it become reasonably clear that health, cumulative, or socioeconomic impacts could be considered in individual licensing proceedings under the Table S-3 Rule.169 Even with the amendment, in fact, there could be some doubt as to whether or not cumulative or socioeconomic impacts were still precluded from consideration. As stated above, the amended Rule and its Statement of Consideration referred explicitly to health effects, but not to socioeconomic or cumulative effects.170 Nevertheless, it did replace the language that had effectively limited licensing boards’ consideration and disclosure of the environmental effects of fuel-cycle activities to the values listed in Table S-3. The new language precluded consideration of the effects “addressed by the Table.”171 Because socioeconomic and cumulative effects were not addressed by the Table, the terms of the amended Rule did not prevent licensing boards from considering such effects and should not have prevented such effects from being disclosed in environmental impact statements. Thus, the amendment of the interim Rule eliminated the NEPA violation.
In view of the foregoing, we hold that the original Table S-3 Rule and the interim Table S-3 Rule, prior to amendment, were invalid. Although the amended interim Rule, and later the final Rule, allowed consideration of health, socioeconomic, and cumulative effects, a number of licenses issued under the original and unamended interim Rules have been challenged. Several of those cases are currently pending in this circuit awaiting this decision.172
D. Economic Feasibility
Petitioner, the State of New York, and Intervenor, the State of Wisconsin, argue that the effluent-release values listed in Table S-3 assume the use of technology that is economically infeasible, even if it is technologically feasible. They argue, first, that the Commission applied an improper standard of economic feasibility in concluding that its projected releases are reasonably foreseeable; second, that the Commission’s cost estimates are incorrect, and third, that the Commission failed to explain the basis of its finding of feasibility.173 We conclude that, although the Commission could have been clearer in setting out its reasons for determining that Table S-3’s predictions are economically feasible, the determination itself was not arbitrary or capricious.174 Therefore, we affirm the Commission on this issue.
*41As Judge Tamm stated in our initial decision in this case, “[t]he Commission should be able to supply the court with a statement of the methods by which its staff arrived at the figures embodied in Table S-3 and by which [its staff] concluded that the waste storage problem is already technologically and economically soluble.”175 The values in Table S-3 are based on assumptions concerning the types of technology that will be used in waste-management and disposal activities.176 For those values to represent reasonably foreseeable environmental effects, as required by NEPA, the use of that technology must be reasonably foreseeable. It must, therefore, be both technologically and economically feasible.177
In concluding that the facilities required to meet the effluent projections of Table S-3 are economically feasible, the Commission applied a standard of whether the facilities are “prohibitively” or “outrageously expensive.”178 Although the Commission’s articulation of the standard is somewhat nebulous, we interpret it to mean that a facility would be economically feasible if its expected cost is no more than the nuclear power industry will pay, under the regulation of the NRC or with the aid of reasonably foreseeable public subsidization.179
New York and Wisconsin argue that “[ejconomic feasibility must be determined in the context of private enterprise, and the question is whether the costs are feasible for the profit-making sector.”180 The basis of their argument seems to be that if it is very expensive to limit effluent discharges to the levels predicted in the Table, the nuclear power industry may economize, at the expense of the public, by allowing higher discharges.181 We find this approach untenable. The standard of economic feasibility may be based on realistic, conservative, and reasoned forecasts of who will have to pay the costs of waste management and disposal, and under what type of compulsion.182 The history of the federal government’s commitment to regulate and subsidize the nuclear power industry, and to operate certain nuclear facilities itself, is clear and long-standing. Moreover, at least the final repository envisioned by the Commission is expected to be developed and operated by the federal government.183 The Commission, in making its prediction of economic feasibility, therefore, was justified in making the reasonable assumption *42that the regulatory structure of the nuclear power industry would remain basically unchanged.184 Under that structure, the NRC will have the power to regulate the industry in such a manner that no private firm will be able to, let alone have an incentive to, cut its costs by allowing the release of more effluents than those projected in Table S-3. If it turns out that the industry cannot pay the full cost of managing and disposing of the waste it has created, it is reasonable to expect the federal government to help out, for society as a whole can be expected to prevent private industry from taking chances in handling nuclear waste. Because the possibility that a firm will have in a socially irresponsible manner is well within the control of the Commission, we hold that the Commission correctly determined that the standard of economic feasibility is that level of costs that the industry in combination with the federal government can reasonably be expected to pay. Of course, we can only hope that the Commission, in fact, exercises its control over the industry in a manner consistent with the public interest.185
New York and Wisconsin challenge, as arbitrary and capricious, both the cost estimates that the Commission accepted,186 and the conclusion that such costs would be feasible.187 The major elements of the Commission’s cost estimates that New York and Wisconsin challenge are the discount rate used and the estimate of decommissioning costs.
In preparing the cost estimates in question, the Commission staff discounted future expenditures to present values at a ten percent rate. In the rulemaking proceeding and again before this court, New York objected to both the discounting procedure and the discount rate used, arguing that a discount rate from zero to two percent would be more appropriate.188 We find, however, that although this argument may be meritorious, it does not undermine the Commission’s finding of feasibility, because both the Hearing Board and the Commission used discount rates of zero and two percent in estimating the range of expected costs.189
New York and Wisconsin also argued before the Commission, and now before this court, that the Commission underestimated, by a factor of ten, the cost of decontaminating and decommissioning (D&D) a reactor.190 Although the Hearing Board took evidence on D&D costs, the Commission ultimately decided that the consideration of *43such costs should be relegated to individual licensing proceedings.191 D&D costs were consequently held to be irrelevant to the economic feasibility of the Table S-3 values. The effluent values in Table S-3 nevertheless continue to include effluents emitted from decommissioned plants.192 We must assume, therefore, that licensing boards will use Table S-3’s values attributable to decommissioning only if they are assured that the decommissioning method assumed by the Table is reasonably likely to be the method used for the individual reactor under consideration. If not, we assume that the licensing board will be free to consider the full environmental costs of whatever decontamination and decommissioning method is expected to be used.193 Thus to this limited extent, we interpret the Table S-3 Rule to be nonpreemptive. Therefore, we reject New York’s and Wisconsin’s contention that Table S-3 is invalid because of inaccuracies in the D&D cost estimates.
Using a variety of discount rates and including the staff’s estimated decommissioning costs,194 the Commission found that the projected waste-management and disposal facilities would entail capital costs of between $71 million and $76 million,195 and operating costs ranging from 0.4 mills/KWh to 1.4 mills/KWh for the once-through cycle and 1.9 mills/KWh to 3.9 mills/KWh for the uranium-only recycle option.196 The Commission observed that such costs represented less than 10% of the capital costs and less than 5% of the operating costs of a typical reactor.197 The Commission found these costs to be feasible simply because they represent only a small fraction of the total cost of building and operating a reactor.198 New York and Wisconsin argue that the finding is arbitrary and capricious because the Commission did not explain why, even under the Commission’s own standard, these costs are feasible. We conclude that, although the Commission’s reasoning is far from transparent, its finding of economic feasibility is neither arbitrary nor capricious. We recognize that an agency must provide a rational connection between facts found and conclusions reached, and a court may not supply a reasoned basis for an agency’s action.199 Nonetheless, a court must uphold a decision of less than ideal clarity if the agency’s path may rea*44sonably be discerned; 200 and in reviewing an agency’s decision, a court can only ensure that the agency properly considered all relevant factors, and that its conclusion does not represent a clear error in judgment.201 In this case, it is apparent that by comparing the projected costs of waste management and disposal to current reactor costs, the Commission was implicitly considering the level of private and, if necessary, public resources that could reasonably be expected to be available for these activities.202 Because these waste-management and disposal expenses represent a comparatively small addition to the resources needed to build and operate a reactor, and because a plant need not be built if it appears that those expenses will render the plant unprofitable, we cannot say that the Commission clearly erred in determining that sufficient resources would be available. The Commission has thus addressed the relevant factor of the availability of private and public resources and reached a conclusion that is within the range of reasonability. We conclude that the Commission’s finding of economic feasibility was not arbitrary or capricious, and dismiss the petition in No. 79-2110.
CONCLUSION
For the foregoing reasons, we hold that the original, interim and final Table S-3 Rules are invalid due to their failure to allow for proper consideration of the uncertainties that underlie the assumption that solidified high-level and transuranic wastes will not affect the environment once they are sealed in a permanent repository. We also hold that the original Rule and the interim Rule, prior to its amendment, are invalid due to their failure to allow for proper consideration of the health, socioeconomic and cumulative effects of fuel-cycle activities. We conclude, however, that the Commission’s finding of economic feasibility was not arbitrary or capricious. We, therefore, vacate all three Rules. Licenses already granted under the Rules are not at issue in this action, and we accordingly express no view as to their validity. The validity of those licenses will be determined in subsequent proceedings.203
Vacated and Remanded.
. The Nuclear Regulatory Commission (NRC) was established by the Energy Reorganization Act of 1974, 42 U.S.C. §§ 5801 et seq. (1976), to assume the licensing and regulatory functions of the Atomic Energy Commission (AEC). The administrative proceedings challenged in this action began under the AEC and continued under the NRC. Hereinafter both agencies will be referred to as the “Commission” or the “NRC.”
. 42 U.S.C. § 4332 (1976).
. “Reprocessing” is a process by which reusable elements of spent fuel are extracted for recycling and the remaining radioactive wastes are concentrated.
. As we stated in our initial opinion in this case, waste “disposal” is a misnomer for what, under currently projected technology, should be termed long-term waste storage. It refers to the last stage of the waste-management process, where wastes are expected to be contained for the tens of thousands of years necessary for them to decay naturally into nontoxic substances. But see p. 480 infra (other disposal methods under consideration).
. 39 Fed.Reg. 14188 (1974).
. 42 Fed.Reg. 13803 (1977).
. 44 Fed.Reg. 45362 (1979) (codified at 10 C.F.R. §§ 51.20(e) & 51.23(c) (1981)). The original Rule was codified at 10 C.F.R. Part 50 App. D and then transferred to 10 C.F.R. §§ 51.20(e) & 51.23(c). 39 Fed.Reg. 26279 (1974). The interim and final Rules amended those two sections. Hereinafter only Federal Register citations will be noted.
Although the original and interim Rules have been superceded by the final Rule, their validity is still at issue. Individual licenses that were granted under those Rules have been challenged in separate actions, many of which are being held in abeyance pending the resolution of the broader issues presented in this case. In this circuit alone there are at least five such cases. Lloyd Harbor Study Group, Inc. v. NRC, No. 73-2266; Aeschliman v. NRC, No. 73-1776; Saginaw Valley Study Group v. NRC, No. 73-1867; NRDC v. NRC, No. 74-1385; Coalition for the Environment v. NRC, No. 77-1905.
.The Rule is referred to as the “Table S-3 Rule” because the Table, which originated in a staff report, was initially labeled “S-3.” See p. 469 infra. In most relevant respects, differences among the three versions of the Table S-3 Rule are insignificant. References to the “Table” or the “Rule,” therefore, refer to all versions.
The fuel cycle activities addressed by the Table include uranium mining and milling, the production of uranium hexafluoride, isotopic enrichment, fuel fabrication, spent-fuel storage and disposal, reprocessing of irradiated fuel, transportation of radioactive materials and management of low-level and high-level wastes. For an explanation of the fuel cycle, see NRDC v. NRC, 547 F.2d 633, 637 n.3 (D.C.Cir.1976), rev’d sub nom. Vermont Yankee Nuclear Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978). As stated above, these cases concern only the waste-management and disposal aspects of the Table. The final version of the Table is reproduced below.
Table S-3. — Table of Uranium Fuel Cycle Environmental Data1
Environmental considerations
Total
Maximum effect per annual fuel requirement or reference reactor year of model 1,000 MWe LWR
Natural Resources Use
Land (acres):
Temporarily committed 2______________ 100
Undisturbed area__________________ 79
Disturbed area____________________ 22 Equivalent to a 110 MWe coal-fired power plant.
*14
Overburden moved (millions of MT)_____________ 2.8Equivalent to 95 MWe coal-fired power plant.
Water (millions of gallons): Discharged to air______________ 160 =2 percent of model 1,000 MWe LWR with cooling tower.
Discharged to water bodies______ 11,090
Discharged to ground----------127
Total______________________ 11,377 4 percent of model 1,000 MWe LWR with once-through cooling.
Fossil fuel:
Electrical energy (thousands of MW-hour)------323 5 percent of model 1,000 MWe LWR output.
Equivalent coal (thousands of MT)___________ 118 Equivalent to the consumption of a 45 MWe coal-fired power plant.
Natural gas (millions of scf)______________ 135 0.4 percent of model 1,000 MWe energy output.
Effluents — Chemical (MT)
Gases (including entrainment):3
SOx_________^............... 4,400
N0X4---------------------1,190 Equivalent to emissions from 45 MWe coal-fired plant for a year.
Hydrocarbons________________ 14
CO_________________________ ' 29.6
Particulates__________________ 1,154
Other gases: F.......................... .67 Principally from UF6 production, enrichment, and reprocessing. Concentration within range of state standards — below level that has effects on human health.
HC1 ... .014
Liquids:
S0-4 ... 9.9From enrichment, fuel fabrication, and
no-3... 25.8 reprocessing steps. Components that con-
Fluoride 12.9 stitute a potential for adverse environmen-
Ca++ _ 5.4 tal effect are present in dilute concentra-
Cl-____ 8.5 tions and receive additional dilution by
Na+ .. 12.1 receiving bodies of water to levels below
NH3 ... 10.0 permissible standards. The constituents
Fe..... .4 that require dilution and the flow of dilution water are:
NH3 — 600 cfs.
N03 — 20 cfs.
Fluoride — 70 cfs.
Tailings solutions (thousands of MT) 240 From mills only — no significant effluents to environment.
Solids____________ 91,000 Principally from mills — no significant effluents to environment.
*15
Gases (including entrainment):
Rn-222 ----------------------------------- Presently under reconsideration by the Commission.
Ra-226 ____________________________ .02
Th-230 ____________________________ .02
Uranium__________________________ .034
Tritium (thousands)__________________ 18.1
C-14______________________________ 24
Kr-85 (thousands)___________________ 400
Ru-106---------------------------- .14 Principally from fuel reprocessing plants.
1-129_____________________________ 1.3
1-131_____________________________ .83
Tc-99------------------------------------ Presently under consideration by the Commission.
Fission products andtransuranics___________________ .203
Liquids:
Uranium and daughters_______________ 2.1 Principally from milling — included tailings liquor and returned to ground — no effluents, therefore, no effect on environment.
Ra-226 ____________________________ .0034 From UF6 production.
Th-230 ____________________________ .0015
Th-234 _______ .01 From fuel fabrication plants — concentration 10 percent of 10 CFR 20 for total processing 26 annual fuel requirements for model LWR.
Fission and activation products_________________5.9 X 10-6
Solids (burned on site):
Other than high level (shallow)___________________________ 11,300 9,110 Ci comes from low level reactor wastes and 1,500 Ci comes from reactor decontamination and decommissioning — buried at land burial facilities. 600 Ci comes from mills — included in tailings returned to ground. Approximately 60 Ci comes from conversion and spent fuel storage. No significant effluent to the environment.
TRU and HLW (deep)________________1.1 X 10 7 Buried at Federal Repository.
Effluents — Thermal (billions of British thermal units)________________ 4,063 5 percent of model 1,000 MWe LWR.
Transportation (person-rem):
Exposure of workers and general public_____________________ 2.5
Occupational exposure (person-rem)______________________ 22.6 From reprocessing and waste management.
1 In some cases where no entry appears it is clear from the background documents that the matter was addressed and that, in effect, the Table should be read as if a specific zero entry had been made. However, there are other areas that are not addressed at all in the Table. Table S-3 does not include health effects from the effluents described in the Table, or estimates of releases of Radon-222 from the uranium fuel cycle or *16estimates ot Technetium-99 released from waste management or reprocessing activities. These issues may be the subject of litigation in the individual licensing proceedings.
Data supporting this table are given in the “Environmental Survey of the Uranium Fuel Cycle,” WASH-1248, April 1974; the “Environmental Survey of the Reprocessing and Waste Management Portion of the LWR Fuel Cycle,” NUREG-0116 (Supp. 1 to WASH-1248); the “Public Comments and Task Force Responses Regarding the Environmental Survey of the Reprocessing and Waste Management Portions of the LWR Fuel Cycle,” NUREG-0216 (Supp. 2 to WASH-1248); and in the record of the final rulemaking pertaining to Uranium Fuel Cycle Impacts from Spent Fuel Reprocessing and Radioactive Waste Management, Docket RM-50-3. The contributions from reprocessing, waste management and transportation of wastes are maximized for either of the two fuel cycles (uranium only and no recycle). The contribution from transportation excludes transportation of cold fuel to a reactor and of irradiated fuel and radioactive wastes from a reactor which are considered in Table S — 4 of § 51.20(g). The contributions from the other steps of the fuel cycle are given in columns A-E of Table S-3A of WASH-1248.
2 The contributions to temporarily committed land from reprocessing are not prorated over 30 years, since the complete temporary impact accrues regardless of whether the plant services one reactor for one year or 57 reactors for 30 years.
3 Estimated effluents based upon combustion of equivalent coal for power generation.
4 1.2 percent from natural gas use and process.
10 C.F.R. § 51.20(e) (1981).
. In this opinion, “EIS” shall refer to the applicant’s Environmental Report as well as the Draft and Final Statements for both the construction and operating license stages. 10 C.F.R. §§ 51.20-51.26 (1981). The Rule specifies a single set of data to be used in all of these documents. Id. at §§ 51.20(e), 51.23(c), 51.-26(a) (1981).
. Licensing and Regulatory Policy and Procedures for Environmental Protection, 10 C.F.R. Part 51 (1981).
Section 102(2)(C) of NEPA requires the EIS to accompany the license application through the existing review process. The Commission’s general rules implementing NEPA provide that when matters covered by the EIS are at issue, the staff will offer the EIS in evidence at the adjudicatory hearing, and any party may take a position and offer evidence as well. The presiding officer will then decide the matters in controversy among the parties. 10 C.F.R. § 51.52(b) (1981). If such matters are resolved by rule, they are generally not subject to attack by way of discovery, proof, argument or other means in the adjudicatory proceeding. See 10 C.F.R. § 2.758 (1981).
. See NRDC v. NRC, 547 F.2d 633, 638-39, 651-52 & n.54 (D.C.Cir.1976), rev’d sub nom. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978); note 14 infra. High-level wastes, which are highly radioactive, are produced in liquid form when spent fuel is reprocessed. Transuranic wastes are nuclides heavier than uranium that are produced in the reactor fuel. These nuclides, which are also highly toxic and may have half-lives of tens of thousands of years, can contaminate pieces of the fuel apparatus. As a result, hardware, particularly from reprocessing plants will have to be disposed of in a permanent repository.
The Commission has considered three possible fuel cycles: the “once-through,” cycle in which fuel is not reprocessed; the uranium-only reprocessing cycle; and, for a short time, the uranium and plutonium reprocessing cycle. Transuranic and high-level wastes are produced in both reprocessing cycles. The uranium-*17only cycle also produces plutonium, which must be disposed of with the transuranic and high-level wastes. In the once-through cycle, the spent-fuel assemblies themselves constitute the waste. Table S-3 is intended to represent the environmental impact of both the once-through and the uranium-only recycle options, which are the only options under current consideration. The Table does so by listing, for each effluent, the highest of the two releases that would be expected under each cycle. 41 Fed.Reg. 45849, 45850 (1977).
. Actually, high-level and transuranic wastes are not even listed in the original Table. See 39 Fed.Reg. 14188, 14191 (1974).
. The final Rule has eliminated this defect by explicitly requiring licensing boards to consider these impacts in individual proceedings pending the promulgation of a more complete generic resolution. 44 Fed.Reg. 45362, 45371 (1979). Petitioners, therefore, only challenge the original and interim Rules on this ground. NRDC Brief at 18, 31, 38.
. Neither the toxicity of radioactive wastes nor the time required for their decay to safe levels can be simply stated. Both involve uncertainties and disputed assumptions and both depend on other variables. See National Academy of Sciences, National Research Council Committee on the Biological Effects of Ionizing Radiation, The Effects on Populations of Exposure to Low Levels of Ionizing Radiation (1980) [hereinafter cited as BEIR Report]; U. S. Nuclear Regulatory Comm’n, Environmental Survey of the Reprocessing and Waste Management Portions of the LWR Fuel Cycle 4-90 to 4-92 (NUREG-0116, Oct. 1976) [hereinafter cited as NUREG-0116], NUREG-0116 is included in the Joint Appendix Supplement as Volume V [hereinafter the Joint Appendix is cited as JA, and the Joint Appendix Supplement as JAS], [To avoid repetition, page citations will refer to NUREG-0116 and not to same page numbers in the JAS.] One federal government report summarizes the problem as follows:
Waste consists of radioactive species of almost all chemical elements; some contain naturally occurring radioactive materials and others contain man-made radioactive materials; the wastes exist as gases, liquids, and solids. Yet for all their variety, radioactive wastes have one thing in common: as long as they remain highly radioactive, they will be potentially hazardous. This potential hazard results from the fact that exposure to and/or uptake of radioactive material can cause biological damage.
In man, it can lead to death directly through intense exposure and a variety of diseases, including cancer, which can be fatal. In addition, radioactive material can be mutagenic thereby transmitting biological damage into the future.
*18The central scientific fact about radioactive material is that there is no method of altering the period of time in which a particular species remains radioactive, and thereby potentially toxic and hazardous without changing that species. Only with time will the material decay to a stable (non-radioactive) element. The pertinent decay times vary from hundreds of years for the bulk of the fission products to millions of years for certain of the actinide elements and long-lived fission products. Thus, if present and future generations are to be protected from potential biological damage, a way must be provided either to isolate waste from the biosphere for long periods of time, to remove it entirely from the earth, or to transform it into non-radioactive elements.
Interagency Review Group on Nuclear Waste Management, Report to the President 8-9 (TID-29442, March 1979) [hereinafter cited as IRG Report].
. NUREG -0116, supra note 14, at 4-72. See also 44 Fed.Reg. 45362, 45374 (1979) (Separate Views of Commissioner Gilinsky).
. See pp. 480-481 infra. See also 44 Fed. Reg. 45362, 45368 (1979); id. at 45374 (Separate Views of Commissioner Gilinsky).
. The term “license” as used in this opinion, shall refer to both construction permits and operating licenses.
. Order Determining Objections to Interrogatories Submitted by Natural Resources Defense Council, In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), Docket No. 50-271, U. S. Atomic Energy Comm’n, Atomic Safety and Licensing Board (May 11, 1971).
. A proposed nuclear plant goes through an extensive licensing process before coming on line. See 10 C.F.R. § 2.50 (1981). Before construction may. begin, the Commission must issue a plant a construction permit. Id. at § 50.10(b). Then, before a plant may begin operation, the Commission must issue it an operating license. Id. at § 50.10(a). At each point, an applicant must prove that it will be able to run the plant properly, that the plant will meet the necessary safety standards, and that the benefits of the plant outweigh its costs. Id. at §§ 50.40, 50.42. Each decision is first rendered by the Atomic Safety and Licensing Board (ASLB) and is then reviewable by the Atomic Safety and Licensing Appeal Board (ASLAB). Id. at § 2.785. The Commission itself has the discretionary power to hear appeals from the ASLAB. Id. at § 2.786. The Commission also has generic rulemaking power. Id. at §§ 2.800-2.801.
. Order Determining Objections to Interrogatories submitted by Natural Resources Defense Council, In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), Docket No. 50-271, U. S. Atomic Energy Comm’n, Atomic Safety and Licensing Board (May 11, 1971). See In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), ALAB-56, 4 A.E.C. 930 (1972).
. 4 A.E.C. at 934 — 36. The Atomic Safety Licensing and Appeal Board did reverse the Licensing Board to the extent that the latter board had excluded evidence on the environmental effects of transporting spent fuel and other wastes to and from the plant, and of decontaminating and decommissioning the plant at the end of its useful life. Id. at 939.
. 37 Fed.Reg. 24191 (1972).
. Id. at 24192. The Commission staff had recently completed a study that reached that conclusion. Environmental Survey of the Nuclear Fuel Cycle (November 6, 1972). The Commission later revised the Survey and published it as Environmental Survey of the Uranium Fuel Cycle (WASH-1248, April, 1974) [hereinafter cited as WASH-1248]. [All references to the Survey will be to WASH-1248, which is included in the Joint Appendix as Volume II. To avoid repetition, page citations will refer to WASH-1248 and not to the same page numbers in the J.A.] The revised Survey added a discussion of waste management and reprocessing based on the testimony at the rulemaking hearing of Dr. F. K. Pittman, Director of the AEC’s Division of Waste Management. WASH-1248 at iv.
. 37 Fed.Reg. 24192-93 (1972).
. The notice referred to WASH-1248, supra note 23, as providing support for Table S-3. 37 Fed.Reg. 24192 n.l.
. WASH-1248 states:
The facility will be designed to prevent the release of significant amounts of radioactive material to the environment under all credible environmental conditions and human actions. Therefore, such wastes will not be released as effluents to the environment.
Id. at S-23.
. 39 Fed.Reg. 14188 (1974).
. Id. at 14191.
. In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), LPB-73-8, 6 A.E.C. 130 (1973).
. In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), ALAB-179, 7 A.E.C. 159, 163 (1974).
. NRDC v. NRC, 547 F.2d 633 (D.C.Cir.1976), rev’d sub nom. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978).
. 547 F.2d at 653-54.
. 547 F.2d at 647-52.
. 547 F.2d at 641. On October 8, 1976, this court stayed its mandate in these actions. Due to the Supreme Court’s decision, Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978), the mandate never issued.
. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978).
. 435 U.S. at 541-42, 98 S.Ct. at 1210-11.
. 435 U.S. at 542-48, 98 S.Ct. at 1210-14.
. 435 U.S. at 549, 98 S.Ct. at 1214. The Supreme Court did not disturb this court’s ruling that, in the absence of a valid generic rule, the environmental impact of the fuel cycle must be dealt with in individual licensing proceedings. 435 U.S. at 538-39, 98 S.Ct. at 1208-09. The Supreme Court also remanded our decision holding Vermont Yankee’s operating license invalid. By the time of the Supreme Court’s decision, that license had been reconsidered in light of the interim Rule. See TAN 480-490 infra. 435 U.S. at 535 n.14, 98 S.Ct. at 1207 n.14; In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), ALAB-421, 6 N.R.C. 25 (1977). The validity of Vermont Yankee’s license is the subject of No. 74-1385, which was originally consolidated with No. 74-1586, but is now being held in abeyance pending decision in the present cases. See Order of Consolidation for Nos. 74-1586, 77-1448, 79-2110, and 79-2131, Feb. 11, 1980; Order Clarifying Application of Order of February 11, 1980 to No. 74-1385, March 12, 1980.
. 435 U.S. at 535 n.14, 98 S.Ct. at 1207 n.14. The parties to the present action have stipulated that this court “may consider the entire record in deciding the challenge to any of the orders under review.” Stipulation filed February 29, 1980.
. 41 Fed.Reg. 45849 (1976).
. NUREG-0116, supra note 14.
. 41 Fed.Reg. 45849 (1976). A table entitled “Summary of Impacts of Reprocessing and Waste Management Per RRY [reference reactor year],” was reproduced in the notice. Id. at 45852 (1976). That table disaggregated Table S-3’s list of environmental impacts, indicating the contribution of each stage of the back end of the fuel to the total impacts. It, therefore, represented a more detailed version of a revised Table S-3.
. See 41 Fed.Reg. 34707, 34709 (1976).
. In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station); CLI-76-18, 4 N.R.C. 471 (1976).
. 41 Fed.Reg. 45849, 45850-51 (1976).
. See note 8 supra (sample Table).
. NUREG-0116, supra note 14, at 2-31 n.“f”, 2-33.
. U.S. Nuclear Regulatory Comm’n, Public Comments and Task Force Responses Regarding the Environmental Survey of the Reprocessing and Waste Management Portions of the LWR Fuel Cycle (NUREG-0116) B-96, B-102 (NUREG-0216, March 1977) [hereinafter cited as NUREG-0216],
. Id at B-107.
. Id. at 3-5 to 3-7; 42 Fed.Reg. 13803 (1977).
. 42 Fed.Reg. 13803 (1977).
. See id. at 13807.
. Id. at 13806.
. 43 Fed.Reg. 15613, 15617 (1978). This amendment was foreshadowed, to a limited degree, by a January 1977 Appeal Board ruling that, in an individual licensing proceeding, the discussion of health effects, such as that previously requested by the NRDC, was required when comparing a proposed nuclear power plant to the alternative of a coal-burning plant. In re Tennessee Valley Authority (Hartsville Nuclear Plant), ALAB-367, 5 N.R.C. 92 (1977).
. In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), CLI -77-10, 5 N.R.C. 717 (1977). Initially, however, in promulgating the interim Rule, the Commission had simply terminated the Licensing Board’s reconsideration of Vermont Yankee’s license. 42 Fed.Reg. 13806 (1976).
. In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), ALAB -421, 6 N.R.C. 25 (1977).
When Vermont Yankee and other facilities were licensed, it was anticipated that spent reactor fuel would be stored on site for only a brief period to allow cooling; thereafter, it would be shipped off site for reprocessing or permanent disposal. Plans for off-site reprocessing and storage had not materialized, however, and in 1977 President Carter suspended indefinitely all commercial reprocessing. 13 Weekly Comp, of Pres. Doc. 506 (1977). Vermont Yankee’s on-site storage capacity of 600 fuel assemblies would have been exhausted by 1977 — and the plant forced to shut down in August 1978 — had the utility not sought and received a license amendment from the Commission allowing it to increase its spent-fuel storage capacity to 2,000 assemblies. In re Vermont Yankee Nuclear Power Corp.. (Vermont Yankee Nuclear Power Station), LBP-77-54, 6 N.R.C. 436 (1977). That license amendment would allow the plant to operate through 1987 without shipping waste to an off-site storage facility.
Intervenors protested that such a license amendment requires a formal EIS, but both the Licensing Board and the Appeal Board ruled that the amendment was “not a major Federal action that significantly affects the quality of the human environment.” 6 N.R.C. at 445; In Re Northern States Power Co. (Prairie Island Nuclear Generating Plant), ALAB-455, 7 N.R.C. 41 (1978). The boards relied on the Commission’s prior denial of NRDC’s petition for rulemaking on the issue of waste disposal, see note 60 infra, as support for their finding that the Commission had “reasonable confidence” that waste disposal facilities would be available when needed. 7 N.R.C. at 49-51. On review of the Appeal Board decision, this court ruled that under the AEA, the Commission *23could make a finding of “reasonable confidence” only following a fact-finding proceeding. Minnesota v. NRC, 602 F.2d 412 (D.C.Cir.1979). The court remanded the case to the Commission for it to reconsider its reasonable confidence in light of the ongoing Table S-3 rulemaking proceedings. In response, the Commission instituted a separate “Waste Confidence" proceeding. 44 Fed.Reg. 61372 (1979).
. In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), ALAB-392, 5 N.R.C. 759, 765 n.7 (1977); see also In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), ALAB-421, 6 N.R.C. 25, 30-32 (Farrar concurring).
. 5 N.R.C. at 765 & n.7.
. 6 N.R.C. at 30-32 (Farrar concurring).
. NRDC petitioned for review of the interim Rule on May 13, 1977. On July 5, 1977, however, NRDC filed a Motion To Hold Petition For Review In Abeyance, which was granted. That motion was based on several events that could have materially altered the facts and law of the case, including the presidential decision of April 7, 1977, see note 56 supra, to defer indefinitely the reprocessing of spent fuel, and the then pending Supreme Court review of this court’s previous decision. NRDC v. NRC, 547 F.2d 633 (D.C.Cir.1976), rev’d sub nom. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978).
At about the time that the interim Rule was under consideration, the NRDC petitioned the Commission to conduct a rulemaking on a related issue: whether the wastes from nuclear power reactors could be permanently disposed of without undue risk to public health and safety. NRDC argued that, until such a determination is made, the Commission has no authority under the Atomic Energy Act (AEA) to issue additional power-plant licenses. The Commission denied the petition without a hearing, ruling that the AEA did not make assured safe disposal a prerequisite for plant licensings. 42 Fed.Reg. 34391 (1977). The Commission contended that the AEA only required safe interim storage and reasonable confidence in an eventual permanent solution. The Commission cited the contemporaneous Table S-3 rulemaking, as well as work by the Energy Research and Development Administration — the agency responsible for waste management — as providing the confidence needed to continue licensing. Id. at 34393. The Second Circuit upheld the Commission’s reading of the AEA, although it expressly distinguished the broad delegation under that statute from the more explicit requirements of NEPA. NRDC v. NRC, 582 F.2d 166, 172 (2d Cir. 1978). Cf. Minnesota v. NRC, 602 F.2d 412 (D.C.Cir.1979) (fact-finding proceeding necessary prior to reaching “reasonable confidence” conclusion).
. 42 Fed.Reg. 26987 (1977).
. See 44 Fed.Reg. 45362, 45366-67 (1979).
. Closing Statement of the Natural Resources Defense Council, June 26, 1978, at 2-5, V JAS at 1364-67.
. First Round of Suggested Staff Questions and Comments on Scope of Proceeding, October 31, 1977, at 6-7, II JAS at 439-40.
. Commission’s Order, February 9, 1978, IV JAS at 1254.
. E.g., Supplementary Testimony of Dr. Irwin C. Bupp, II JAS at 662; Statement of Peter N. Skinner, September 30, 1977, III JAS at 1099.
. E.g., Staff Testimony on Economic Data to Support the Feasibility of the S-3 Model, February 3, 1978, III JAS at 1195; Final Written Statement on Behalf of Commonwealth Edison, June 26, 1978, XI-1 to XI-39, II JAS at 763-801.
. 44 Fed.Reg. 45362 (1979).
. See note 8 supra; 44 Fed.Reg. 45362, 45369, 45373 (1979).
. The Commission stated:
The rulemaking record makes clear that effluent release values, standing alone, do not meaningfully convey the environmental significance of uranium fuel cycle activities. The focus of interest and the ultimate measure of impact for radioactive releases are the resulting radiological dose commitments and associated health effects. To convey in understandable terms the significance of releases in the Table, the Hearing Board recommended that the modified Table be accompanied by an explanatory narrative promulgated as part of the rule. The recommended narrative would also address important fuel cycle impacts now outside the scope of the Table, including socioeconomic and cumulative impacts, where these are appropriate for generic treatment. The Commission has directed the NRC staff to prepare by October 1 such a narrative, as described in more detail later in this notice. The narrative will be submitted for public comment in a further rulemaking.
Id. at 45362.
. Id.
. Id. at 45368 (citing IRG Report, supra note 14, at 42).
. Id. at 45369 (footnote omitted).
. Id. at 45367. The Commission also agreed with the Hearing Board that the costs of decontamination were site specific and not within the Rule. Id. at 45367 n. 18.
. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 549, 98 S.Ct. 1197, 1214, 55 L.Ed.2d 460 (1978). We are here reviewing the original, interim and final Rules on their faces — not the procedures that produced them. We also do not consider whether nuclear power plants are “good” or “bad.” Congress has left it up to the Commission to determine whether, in individual instances, the benefits of nuclear power outweigh the costs. All we consider is whether the Commission has complied with the mandates of the APA and NEPA in performing this task.
. 5 U.S.C. § 706(2)(A) (1976).
. See Calvert Cliffs’ Coordinating Committee, Inc. v. AEC, 449 F.2d 1109 (D.C.Cir.1971) (review of rules implementing NEPA to determine whether they complied with NEPA).
. Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 823, 28 L.Ed.2d 136 (1971); Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285-86, 95 S.Ct. 438, 441-42, 42 L.Ed.2d 447 (1974).
. 42 U.S.C. § 4332(2)(C) (1976). See Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 538-39, 98 S.Ct. 1197, 1208-09, 55 L.Ed.2d 460 (1978).
. 42 U.S.C. § 4332(2)(C) (1976).
. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 539, 98 S.Ct. 1197, 1209, 55 L.Ed.2d 460 (1978).
. See, e.g., Trout Unlimited v. Morton, 509 F.2d 1276, 1282 (9th Cir. 1974); Environmental Defense Fund, Inc. v. Corps of Engineers, 325 F.Supp. 749, 759 (E.D.Ark.1971).
. Weinberger v. Catholic Action of Hawaii/Peace Education Project, 454 U.S. 139, 143, 102 S.Ct. 197, 201, 70 L.Ed.2d 298 (1981).
. id.
.Id.
. Andrus v. Sierra Club, 442 U.S. 347, 350, 99 S.Ct. 2335, 2337, 60 L.Ed.2d 943 (1979) (footnote omitted).
. 449 F.2d 1109 (D.C.Cir.1971).
. 449 F.2d at 1118.
. Under NEPA’s Rule of Reason, an agency must consider only “reasonably foreseeable” environmental impacts. See, e.g., Scientists’ Institute for Public Information, Inc. v. AEC, 481 F.2d 1079, 1092 (D.C.Cir.1973); Carolina Environmental Study Group v. United States, 510 F.2d 796, 798 (D.C.Cir.1975).
. 449 F.2d at 1114.
. 449 F.2d at 1114, 1115.
. It is well accepted that courts must exercise heightened review of agencies’ compliance with section 102(2)(C) of NEPA. See NRDC v. SEC, 606 F.2d 1031, 1048 (D.C.Cir.1979); Scientists’ Institute for Public Information, Inc. v. AEC, 481 F.2d 1079, 1094 (D.C.Cir.1973); Calvert Cliffs’ Coordinating Committee, Inc. v. AEC, 449 F.2d 1109, 1114 (D.C.Cir.1971). See also W. Rogers, Environmental Law 716-17 (1977) (procedural requirements of NEPA “enforced with a vengeance”).
. The Commission does not argue that the environmental impact of fuel-cycle activities is insignificant, nor does it appear that such an argument could legitimately be made. The AS-LAB has stated that
although the adverse environmental consequences of the fuel cycle attributable to individual nuclear generating plants are small, they are not without any weight. .. . [l]n a close situation they might throw the cost/benefit balance against issuing a construction permit. This opens the distinct possibility that construction permits would be vulnerable to judicial challenge under NEPA if issued after hearings which did not take relevant and potentially significant adverse environmental factors into account.
In re Potomac Electric Power Co. (Douglas Point Nuclear Generating Station, Units 1 and 2), ALAB-218, 8 A.E.C. 79, 87 (1974). This view was approved by the Commission, which noted that excluding fuel-cycle effects in a close case would place the boards in the almost untenable posture of refusing to consider known relevant environmental factors which could possibly affect the cost-benefit balance. In re Philadelphia Electric Co., et al. (Peach Bottom Power Station, Units 2 and 3), CLI-7432, 8 A.E.C. 217, 220 (1974).
.The issues in Calvert Cliffs’ were presented in a posture similar to that of this case. In that case, as in this, we reviewed regulations that the AEC had promulgated to govern the implementation of NEPA.
. See pp. 470, 472, 474-475 supra; In re Potomac Electric Power Co. (Douglas Point Nuclear Generating Station), ALAB-218, 8 A.E.C. 79, 88-90 (1974); In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Power Station), ALAB-421, 6 N.R.C. 25, 30-32 (1977) (Farrar concurring).
. Contrary to the implication of Judge Wilkey’s dissent, see, e.g., Wilkey op. at 531, 534, 540-545, the NRDC does not challenge, and we do not decide, the reasonableness of the Commission’s conclusion that the probabilities favor the zero-release assumption. Rather, the NRDC objects to the fact that the Commission has not factored the consideration of uncertainties into the licensing process in such a manner that they can potentially affect the outcome of any decision to license a plant.
. In the original Rule’s Statement of Consideration, the Commission made no reference to any uncertainty underlying its conclusion that a permanent repository would be built. In contrast, in the Statements of Consideration for the interim and final Rules, the Commission repeatedly acknowledged the uncertainties involved in its “judgments.” See note> 118 infra. The Commission nonetheless stated that there was “no advantage in having licensing boards repeatedly weigh for themselves the effect of uncertainties on the selection of fuel cycle impacts for use in cost-benefit balancing.” 44 Fed.Reg. 45362, 45369 (1979). The Commission believed that the
“question [was] properly dealt with in this rulemaking as part of choosing what impact values should go into the fuel cycle rule. The Commission concIude[d], [that] having noted that uncertainties exist, that for the limited purpose of the fuel cycle rule it [was] reasonable to base impacts on the assumption which the Commission believefd] the probabilities favor.”
Id. Commissioner Bradford, writing separately, stated that he thought
“that the Commission [had gone] too far in terming its assumption that a ‘bedded salt repository or its equivalent will be found’ to be a ‘judgment’. [Instead,] little more can be said by a prudent regulatory agency at this time in the face of this record and the general uncertainty than that the direction of current federal programs makes a bedded salt repository a responsible working assumption for NEPA purposes.”
Id. at 45373 (Separate Views of Commissioner Bradford). See p. 482 infra.
. Contrary to Judge Wilkey’s assertions, see, e.g., Wilkey op. at 537, our conclusions do not turn on a characterization of the zero-release assumption as a finding of fact. See pp. 481-485 infra.
. See p. 475.
. See, e.g., Izaak Walton League of America v. Marsh, 655 F.2d 346, 377 (D.C.Cir.), cert. denied sub nom. Atchison, Topeka & Santa Fe *29Railway Co. v. Marsh, 454 U.S. 1092, 102 S.Ct. 657, 70 L.Ed.2d 630 (1981); Alaska v. Andrus, 580 F.2d 465, 473 (D.C.Cir.), vacated in part sub nom. Western Oil & Gas Assoc. v. Alaska, 439 U.S. 922, 99 S.Ct. 303, 58 L.Ed.2d 315 (1978); Concerned About Trident v. Rumsfeld, 555 F.2d 817, 830 (D.C.Cir. 1976).
Risk, in this context, refers to the product of the environmental damage that could occur and the probability of its occurrence. Thus, an environmental risk can be high if the probability is high that damage will occur, even if the damage itself would not be terribly severe. Alternatively, an environmental risk can be significant if the probability is low that damage will occur, but the possible damage, should it occur, would be severe. Thus, even if the probability that environmental damage will occur is very low, the risk is nonetheless significant if the potential damage is sufficiently severe. These risks are, therefore, of the same character as other environmental damage costs and, Judge Wilkey’s dissent notwithstanding, Wilkey op. at 536-37, they must be considered in the NEPA inquiry. See, Ethyl Corp. v. Environmental Protection Agency, 541 F.2d 1, 25 n.52 (D.C.Cir.1976) (en banc), cert. denied sub nom. E.I. DuPont de Nemours & Co. v. Environmental Protection Agency, 426 U.S. 941, 96 S.Ct. 2662, 49 L.Ed.2d 394 (all environmental costs involve some degree of uncertainty); Carolina Environmental Study Group v. United States, 510 F.2d 796, 799 (D.C.Cir.1975) (it is necessary “to consider the probabilities as well as the consequences of certain occurrences in assessing [an action’s] environmental impact”).
. When confronting a set of uncertain environmental effects, or an environmental risk that is known, an agency’s goal should be to trace each reasonably foreseeable contingency and determine, first, the likelihood of its occurring, and second, the environmental damage that it would entail should it occur. Of course, precision in this context can only be the ideal, and the detail with which an agency must consider and disclose the likelihood and nature of each contingency is somewhat flexible. If a contingency is not expected to entail a serious environmental risk less detail is required. See, e.g., Izaak Walton League of America v. Marsh, 655 F.2d 346, 377 (D.C.Cir.), cert. denied sub nom. Atchison, Topeka & Santa Fe Railway Co. v. Marsh, 454 U.S. 1092, 102 S.Ct. 657, 70 L.Ed.2d 630 (1981). Alternatively, an agency may evaluate an uncertain environmental effect by considering and disclosing a worst-case scenario along with an estimation of the probability that the scenario will occur. See North Slope Borough v. Andrus, 642 F.2d 589, 605 (D.C.Cir.1980); 40 C.F.R. § 1502.22 (1981). At the outset of the NEPA inquiry, however, a federal agency bears the obligation to seek information concerning the environmental consequences of proposed actions. Alaska v. Andrus, 580 F.2d 465, 473 (D.C.Cir.1978), vacated, in part, sub nom. Western Oil & Gas Assoc. v. Alaska, 439 U.S. 922, 99 S.Ct. 303, 58 L.Ed.2d 315 (1978); Concerned About Trident v. Rumsfeld, 555 F.2d 817 (D.C.Cir.1977).
. See, e.g., Alaska v. Andrus, 580 F.2d 465, 473 n.36 (D.C.Cir.1978), vacated, in part, sub nom. Western Oil & Gas Assoc. v. Alaska, 439 U.S. 922, 99 S.Ct. 303, 58 L.Ed.2d 315 (1978); Scientists’ Institute for Public Information, Inc. v. AEC, 481 F.2d 1079, 1092 (D.C.Cir. 1973).
. Bowman Transportation, Inc. v. Arkansas-Best Freight, Inc., 419 U.S. 281, 285, 95 S.Ct. 438, 441, 42 L.Ed.2d 447 (1974); Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 419, 91 S.Ct. 814, 825, 28 L.Ed.2d 136 (1971).
. WASH-1248, supra note 23.
. Id. at G-6 to G-7.
. Id. at G-7.
. NUREG-0116, supra note 14, at 4 — 71 to 4-73.
. Id. at 4-72.
. Id. at 4-29. The present status of the RSSF is also, at best, uncertain. See Office of Management and Budget, Major Themes & Additional Details, Fiscal Year 1983, at 160 (1982); Vepco Says Crisis Near in Used Nuclear Fuel Storage, Washington Post, March 1, 1982, at Al col. 2.
. Transcript of Hearing at 729-30, II JAS at 622-23.
. IRG Report, supra note 14. This report was entered in the record in draft form and discussed in an oral presentation. See 44 Fed. Reg. 45362, 45368 n.24 (1979).
. IRG Report, supra note 14, at 45.
. Id. at 17.
. Statement of Todd R. LaPorte, Oct. 3, 1977, I JAS at 393; Statement of Dr. Terry Lash, Oct. 3, 1977, II JAS at 437.
. IRG Report, supra note 14, at 87.
. NUREG-0116, supra note 14, at 4-94. See p. 471 supra.
. 44 Fed.Reg. 45362, 45368 (1979) (footnote omitted).
. Id. From the time it proposed the original Rule through its promulgation of the final Rule, the Commission became increasingly candid in its acknowledgement of uncertainties underlying permanent waste disposal. In promulgating the original Rule, the Commission referred to no uncertainty. In promulgating the interim Rule, however, the Commission “observed that there [were] gaps in the information needed for detailed assessment of waste management and disposal technology.” 42 Fed.Reg. 13803, 13805 (1977). Nevertheless, the Commission believed that such uncertainties did not prevent it “from making an informed and reasoned judgment ... regarding the environmental impacts which may flow from waste management and reprocessing activities.” Id. Even more candidly, in the final rulemaking, the Commission stated that it was its “judgment that a suitable bedded-salt repository site or its equivalent will be found, but the Commission note[d] ... that areas of uncertainty remain regarding both the likelihood of finding a site and the probability that it will perform as expected.” 44 Fed.Reg. 45362, 45368 (1979) (footnote omitted). The Commission found that its staffs assumption that there would be no release of effluents from a sealed repository was “afford[ed] a reasonable basis” by evidence that was “tentative but favorable.” Id. The Commission was convinced that “the general evidence [presented], coupled with the absence of any strong argument that a site cannot be found, probably affords as strong a record as can be made on the issue until a specific site has been thoroughly investigated and found to be suitable.” Id. (emphasis added).
. 44 Fed.Reg. 45362, 45369 (1979) (footnote omitted). See note 97 supra.
. See Ecology Action v. AEC, 492 F.2d 998, 1002 (2d Cir.1974); Note, The Use of Generic Rulemaking to Resolve Environmental Issues in Nuclear Power Plant Licensing, 61 Va.L.Rev. 869 (1975).
.We express no view as to whether this type of generic balancing could be done in the context of this case.
. 449 F.2d 1109, 1118 (D.C.Cir.1971). See pp. 476-477 supra.
. At this point, the temporary RSSF has not been built. See note 109 supra. Once it is built, the uncertainty over the development of a permanent repository translates into a possibility that wastes will remain indefinitely at the RSSF. Prior to the point at which the RSSF is built, however, that uncertainty translates into the possibility that the wastes will remain either at reprocessing plants or at the power plants themselves. Similarly, uncertainty concerning the development of the RSSF and of reprocessing plants gives rise to the possibility that wastes will have to remain in individual power plants’ spent-fuel pools. To the extent that any of these possibilities entails significant risks, those risks must be factored into the NEPA calculus. If the risks are generic, they can be assessed, and possibly weighed against generic benefits, in generic rulemaking proceedings. If they are site-specific, however, they must be assessed with particular reference to the relevant site, and balanced along with other site-specific costs against site-specific benefits. Cf. Minnesota v. NRC, 602 F.2d 412 (D.C.Cir.1979) (before allowing expansion of on-site repositories, NRC must examine possibility that spent fuel will remain on site due to lack of temporary or permanent off-site repositories); Storage and Disposal of Nuclear Waste, 44 Fed.Reg. 61373 (1979) (generic rule-making initiated in response to Minnesota).
. See note 93 supra (rule of reason); pp. 479-481 supra (Commission recognition of uncertainties involved).
. Calvert Cliffs’ Coordinating Committee, Inc. v. AEC, 449 F.2d 1109, 1118 (D.C.Cir.1971).
. NRC Brief at 36-38.
. See In re Vermont Yankee Nuclear Power Corp. (Vermont Yankee Nuclear Generating Station), ALAB-421, 6 N.R.C. 25, 30-32 (1977) (Farrar concurring).
. NRC Brief at 37.
. See p. 481 supra.
. See pp. 482^483 supra.
. As stated above, it may be that certain costs can be assessed and balanced against certain benefits at the generic level. See p. 482 supra. That, however, is not what the NRC has done here.
. See p. 482 supra.
. See p. 476 supra. This is not a simple matter of how and where the environmental impact of an action is disclosed as IntervenorRespondent Commonwealth Edison et al. contend. See Commonwealth Edison et al. Brief at 12 n.4; Scientists’ Institute for Public Information, Inc. v. AEC, 481 F.2d 1079, 1092-93 (D.C.Cir.1973).
. We emphatically do not hold that the Commission may not proceed in the face of uncertainty, as Judge Wilkey maintains, see, e.g., Wilkey op. at 540, or that the uncertainty surrounding the zero-release assumption may be factored into the cost-benefit calculus only at the individual licensing level. Judge Wilkey has seriously misconstrued this issue. We do, however, hold that where the Commission has not factored the uncertainties surrounding its assumptions into a generic rule like the Table S- 3 Rule, it cannot also preclude individual licensing boards from considering those uncertainties when applying the values provided by the Rule.
. At one point in its opinion remanding this case, the Supreme Court implied that this court could review the Table S-3 Rule only under the arbitrary and capricious standard. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 535-36 n.14, 98 S.Ct. 1197, 1207-08 n.14, 55 L.Ed.2d 460. Although we do not believe that this was the Supreme Court’s intention — particularly in view of the fact that the later versions of the Rule reveal that the zero-release assumption may not have been simply a factual finding — we reach the same result under the “not in accordance with law” standard and the “arbitrary and capricious” standard.
. Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285, 95 S.Ct. 438, 441, 42 L.Ed.2d 447 (1974); Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 823, 28 L.Ed.2d 136 (1971). See p. 475 supra.
. Judge Wilkey maintains that an agency may assess an uncertain environmental effect generically and prescribe that individual EIS’s reflect only a most probable — or modal — value for the effect. See, e.g., Wilkey op. at 537-539. We strongly disagree. NEPA requires the consideration and disclosure of significant uncertainties concerning environmental effects. A generic rule limiting consideration and disclosure to modal values would fly in the face of that requirement, allowing individual decisions to be made without regard to improbable yet potentially severe environmental damage. Improbable yet potentially severe environmental damage is as much an environmental cost as certain or probable damage, and to allow decisions to be made without consideration of any significant — even if probabilistic — environmental effect constitutes a violation of NEPA. Judge Wilkey states that an agency’s consideration of uncertainties prior to promulgating a generic rule satisfies NEPA’s requirement that uncertainty be considered. Id. In this case, however, that consideration was undertaken in the abstract. It did not allow the agency to reflect upon whether potentially severe environmental costs are warranted by the benefits expected to attend the licensing of a particular plant or nuclear plants in general. That consideration, therefore, was not sufficient to comply with NEPA.
. See NRDC Brief at 31-38.
. See NRC Brief at 34-39.
. See note 8 supra (copy of Table S-3).
. See Conclusions and Recommendations of the Hearing Board (October 26, 1978), IV JAS at 1439-40.
. Id. at 1443-46.
. See pp. 476 — 477 supra.
. See North Slope Borough v. Andrus, 642 F.2d 589, 605 (D.C.Cir.1980).
. These include uncertainties surrounding the health, socioeconomic, and cumulative impacts of the back end of the uranium fuel-cycle and the economics of the projected waste-management and disposal facilities. See pp. 486-494 infra.
. NRDC Brief at 18-31.
. See, e.g., Kleppe v. Sierra Club, 427 U.S. 390, 410, 96 S.Ct. 2718, 2730, 49 L.Ed.2d 576 (1976) (cumulative effects); NRDC v. Callaway, 524 F.2d 79, 88 (2d Cir. 1975) (cumulative effects); Named Individual Members of San Antonio Conservation Society v. Texas Highway Dept., 446 F.2d 1013 (5th Cir. 1971), cert. denied, 406 U.S. 933, 92 S.Ct. 1775, 32 L.Ed.2d 136 (1972) (cumulative effects); Trinity Episcopal School Corp. v. Romney, 523 F.2d 88, 93 (2d Cir. 1975) (socioeconomic effects); Prince George’s County v. Holloway, 404 F.Supp. 1181, 1186 (D.D.C.1975) (socioeconomic effects); 40 C.F.R. §§ 1508.7, 1508.8 (1981) (health, social, economic, cumulative).
. See, e.g., Silva v. Lynn, 482 F.2d 1282, 1285 (1st Cir. 1973); “[The EIS] must be written in language that is understandable to nontechnical minds and yet contain enough scientific reasoning to alert specialists to particular problems within the field of their expertise.” (quoting Environmental Defense Fund, Inc. v. Corp. of Engineers, 348 F.Supp. 916, 933 (W.D.Miss.1972)). Cf. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 557, 98 S.Ct. 1197, 1218, 55 L.Ed.2d 460 (1978) (the reports used to prepare the EIS may be technical in nature); 40 C.F.R. § 1502.8 (1981) (EIS should be written so that “decisionmakers and the public can readily understand them”).
. 40 C.F.R. § 1508.8 (1981). Environmental impacts must be considered and disclosed in these terms in order to facilitate the comparison of alternative actions that entail different types of environmental damage — for example, radiological and chemical releases. Furthermore, an EIS must translate technical effluent data into such terms if it is effectively to disclose the environmental impacts of a proposed project to its intended readership: interested members of the public, public servants, and legislators. That audience cannot be expected to convert curies or mrems into such costs as cancer deaths, or social disturbance. Moreover, the expression of environmental costs in terms of actual health and socioeconomic effects is necessary to convey to that audience the basis of an agency’s decision. The impacts of nuclear reactors on human health and society are matters of legitimate scientific debate, and NEPA gives the public the right to know the tangible environmental costs that the Commission expects it to pay in return for the anticipated benefits of such a plant.
.39 Fed.Reg. 14188, 14191 (1974); 42 Fed. Reg. 13803, 13806 (1977). The staff report supporting the original Rule did not address the health, socioeconomic or cumulative effects of the fuel cycle at all; and the report supporting the interim Rule contained only cursory references to those effects. See, e.g., NUREG-0116, supra note 14, at 2-19 to 2-20, 4 — 168 to 4 — 171.
. See p. 489 infra.
. In re Tennessee Valley Authority (Hartsville Nuclear Plant), ALAB-367, 5 N.R.C, 92 (1977).
. 43 Fed.Reg. 15613, 15617 (1978) (change in footnote 1); see id. at 15616.
. Id. at 15617 (emphasis added).
. 44 Fed.Reg. 45362, 45371 (1979).
. Id.; see 46 Fed.Reg. 15154 (1981) (proposed narrative Rule); note 168 infra.
. NRDC Brief at 18-31.
. 39 Fed.Reg. 14188, 14191 (1974); 42 Fed. Reg. 13803, 13806 (1977) (emphasis added).
. See p. 469 supra.
. In proposing the original Rule, the Commission commented that “such matters, if they are to be considered at all, be considered in a generic fashion through a rulemaking process.” 37 Fed.Reg. 24191, 24192 (1972).
. 44 Fed.Reg. 45362, 45364 (1979).
. Id.
. As is true of the Commission’s treatment of the zero-release assumption, this violation of NEPA could also be characterized as being arbitrary and capricious. See p. 485 & note 135 supra. Under NEPA, the health, socioeconomic and cumulative effects of a proposed action must be considered and disclosed. The presence of such effects is relevant to an agency’s decision to promulgate a generic rule prescribing the scope of its consideration and disclosure of the environmental effects of proposed actions prior to taking such actions. In promulgating the original and interim Table S--3 Rules, however, the Commission ignored the presence of those types of environmental effects. Its action, therefore, was arbitrary and capricious. See p. 475 supra.
. In re Tennessee Valley Authority (Harts-ville Nuclear Plant), ALAB-367, 5 N.R.C. 92 (1977).
. NRC Brief at 41-42.
. In re Tennessee Valley Authority (Harts-ville Nuclear Plant), ALAB-367, 5 N.R.C. 92 (1977).
. Id. at 103 n.52.
. We also cannot accept the argument that the Commission made a factual finding that these fuel cycle effects were insignificant. The Commission has acknowledged:
A record is not yet available to support a comprehensive rule dealing with all generic aspects of fuel cycle impacts relevant to reactor licensing.... The table of impacts adopted as a final rule ... serves as an important first step [by] relieving adjudicatory boards from the need to determine those numerical impacts of the uranium fuel cycle which have been extensively considered in generic rulemaking. Ultimately, however, the impacts of the releases and not the releases themselves dictate the standards the Commission must set.
44 Fed.Reg. 45362, 45363 (1979). The Commission then initiated the process to consider the significance of the health, socioeconomic, and cumulative effects of the uranium fuel cycle in a generic rulemaking proceeding. Id. at 45371; see 46 Fed.Reg. 15154 (1981). The proposed outcome of that proceeding is to find most of these effects to be insignificant. Id. at 15154. That result is, at this stage only proposed. Even if adopted, it is not before us and we express no opinion on whether the Commission may make such a determination by rule, whether an EIS is required for such a determination, or whether such a determination can be adequately supported. Before deciding not to *40delay the issuance of our opinions pending completion of the narrative rulemaking, see Edwards op. at n.2, we ordered the NRC to advise us as to the status of that proceeding. In response, the Commission stated that
The Commission has received public comments regarding the proposed rule, and the NRC staff is continuing to analyze those comments and prepare proposed responses for the Commission’s approval. As yet, however, a proposed final rule has not been submitted to the Commission and no date has been set for such submission. Therefore it appears that the Commission’s present fuel cycle rule (Table S-3) is not likely to be supplanted in the immediate future.
Letter from NRC, dated October 16, 1981.
. See 43 Fed.Reg. 15613 (1978).
. See p. 487 supra.
. See id.
. See note 7 supra.
. New York argued before the Hearing Board that the Commission must address the economic feasibility of its projections. The Commission staff argued, in response, that the issue need not be addressed. II JAS at 439-40. The Hearing Board, however, held that the issue should be addressed and took evidence from New York, Wisconsin, and Ohio purporting to show that the values projected in Table S-3 were not economically feasible. See Commission’s Order, Feb. 9, 1978, IV JAS at 1254 (Commission affirming Hearing Board).
. Judge Wilkey’s dissent questions why we treat this issue differently than the uncertainty issue. The answer is simple: the issues are significantly different. The NRDC challenges not the conclusion that the zero-release assumption is reasonable, but rather the arbitrary exclusion of the uncertainties surrounding that assumption from any consideration in the licensing process. Here, New York challenges the Commission’s conclusion that the facilities will be economically feasible. Economic uncertainties, of course, still must be factored into the analysis. See note 145' supra.
. NRDC v. NRC, 547 F.2d 633, 661 (D.C.Cir.1976) (Tamm, J., concurring), rev’d sub nom. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978) (NEPA requires Commission to find technological and economic feasibility). See also 547 F.2d at 658 (same).
. See, e.g., pp. 480-481 supra.
. In the abstract, economic and technological feasibility may be closely related. A piece of technology that is so expensive that no private or public institution could ever amass the resources necessary to develop it may be characterized as technologically infeasible when in fact it is feasible as a purely technological matter. That type of economic feasibility is addressed, though indirectly, in Sections A and B, above. The concept of economic feasibility that New York and Wisconsin are questioning seems to be one that is well within the bounds of affordability in this extreme sense. The concept of economic feasibility that underlies New York’s and Wisconsin’s argument seems to be one of affordability by particular economic institutions that can reasonably be expected to have responsibility for bearing the cost of a piece of technology.
. 44 Fed.Reg. 45362, 45367 (1979).
. In its Statement of Consideration the NRC stated:
Whether nuclear power is good business is not an issue in this rulemaking. The fuel cycle rule will be used only when somebody has decided, rightly or wrongly, that nuclear power is sufficiently viable economically to warrant applying for a reactor license. Once the reactor has operated, back-end fuel cycle activities must be carried out, whatever the cost.
Id.
. New York/Wisconsin Reply Brief at 6.
. New York/Wisconsin Brief at 28.
. Scientists’ Institute for Public Information, Inc. v. AEC, 481 F.2d 1079, 1092 (D.C.Cir.1973).
. NUREG-0116, supra note 14, at 4-29, 4-71.
. See Sierra Club v. Morton, 510 F.2d 813, 827-28 (5th Cir. 1975) (NEPA analysis does not have to ignore continuing regulatory and monitoring responsibilities); NRDC v. Morton, 458 F.2d 827, 837 (D.C.Cir.1972) (elimination of fundamental legislative schemes, such as antitrust laws, need not be considered).
. In its brief, the NRC states that:
the “operators” of fuel cycle facilities, whether government or private, will be subject to NRC regulation and will not have the option of cutting costs by allowing further releases if these would exceed the Commission’s regulatory standards. The Commission’s finding of economic feasibility thus in a sense expresses agency policy not to lower standards, even if this were legally permissible, on account of the predicted level of costs.
NRC Brief at 27 (emphasis added). A future change in such an agency policy may well be an agency action that significantly affects the environment.
As Judge Edwards correctly points out, an agency may not simply assume that the federal government will subsidize the prevention of foreseeable environmental damage. Nevertheless, in an area such as this, where there is a long history of federal involvement — including subsidization — we cannot say that an agency errs if it projects that a relatively minimal level of federal funding would become available if it turns out that private industry alone cannot prevent serious environmental damage.
. New York/Wisconsin Brief at 35-50.
. See 44 Fed.Reg. 45362, 45367'(1979); Hearing Board Conclusions and Recommendations, at 58, IV JAS at 1480.
. Report of Hearing Board, August 31, 1978, at 126-130, III JAS at 1056-1060.
. 44 Fed.Reg. 45362, 45367 & n.18 (1979); Hearing Board Conclusions and Recommendations, at 52-53, 57-58; IV JAS at 1475-1476, 1479-80.
. Hearing Board Conclusions and Recommendations, at 54-56, IV JAS 1476-78; New York/Wisconsin Brief at 35-50.
. 44 Fed.Reg. 45362, 45367 n.18 (1979); see Hearing Board Conclusions and Recommendations, at 56, IV JAS at 1478.
. See note 8 supra.
. The Hearing Board stated that “[t]he cost [of decommissioning a reactor] is specific to each plant and depends on the method chosen. Controversies regarding the decommissioning costs should be resolved in individual reactor licensing proceedings.” Hearing Board Conclusions and Recommendations, at 56, IV JAS at 1478 (emphasis added). Similarly the Commission found the “costs of decommissioning a power reactor ... [to be] facility-specific and should be considered in individual reactor proceedings rather than included among the costs of the fuel cycle activities which are the subject of the generic rule." 44 Fed.Reg. 45362, 45367 n.18 (1979) (emphasis added).
. Hearing Board Conclusions and Recommendations, at 52, 54, IV JAS 1474, 1476.
. These figures are for the once-through and uranium-only reprocessing cycles, respectively. Hearing Board Conclusions and Recommendations, at 51, IV JAS at 1473.
. These ranges are from zero percent discount rate to ten percent discount rate. Id. at 52, 54, 57, IV JAS at 1474, 1476, 1479.
. These percentages are based on cost estimates, excluding waste management and disposal, of one billion dollars in capital costs and operating expenses of 30 mills/KWh for the model or “typical” reactor. The 5% figure is based on the operating cost of the once-through cycle (1.4 mills/KWh — 30 mills/KWh). It therefore represents an upper bound, because it is assumed that no reprocessing would take place unless the value of uranium that would be recovered makes up for the higher operating cost of the uranium-only reprocessing cycle. Such an assumption is clearly reasonable. Hearing Board Conclusions and Recommendations, at 51 & n.27, 52, 58-59, IV JAS at 1473 & n.27, 1474, 1480-81.
. Id. at 58, IV JAS at 1480; see 44 Fed.Reg. 45362, 45367 (1979).
. SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 1577, 91 L.Ed. 1995 (1947).
. Colorado Interstate Gas Co. v. Fed. Power Comm’n, 324 U.S. 581, 595, 65 S.Ct. 829, 836, 89 L.Ed. 1206 (1945).
. Bowman Transportation, Inc. v. Arkansas-Best Freight System, 419 U.S. 281, 285, 95 S.Ct. 438, 441, 42 L.Ed.2d 447 (1974); Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 823, 28 L.Ed.2d 136 (1971).
. See Hearing Board Conclusions and Rec-. ommendations, at 58 & n.41, IV JAS 1480 & n.41; Transcript at 1291, II JAS 744 (comment of Hearing Board member Briggs).
See also note 185 supra. Our doubts about the Commission judgment that such resources will be available would be greater if these costs did not' represent such a relatively small increment to the present costs of nuclear power.
.See note 7 supra.
.The as yet unadopted Narrative makes clear that the current staff thinking favors deep layers of bedded salt. See generally U.S. Nuclear Regulatory Commission, Environmental Survey of the Reprocessing & Waste Management Portions of the LWR Fuel Cycle, NUREG-0116 (Oct. 1976)'.
. At the adoption of Table S-3 as a final rule, the NRC listed 70 licensed nuclear power plants. See 44 Fed.Reg. 45362, 45370 n.30 (Aug. 2, 1979). That number appears to have changed, however. See note 8 infra and accompanying text.